DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
.
Claims 1, 2, 7–12, 17–22, and 27–30 are rejected under 35 U.S.C. 103 as being unpatentable over Tan in view of Siomina (WO 2014/053995 A1, hereinafter Siomina)
Regarding claim 1 Tan discloses “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
Thus it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
Regarding claim 2, Tan discloses a method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising evaluating configuration changes by”, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Tan in view of Siomina.
Claim 7 depends from claim 1.
“The method of claim 1.”
With respect to “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the quality score represents the quality of user experience associated with the wireless network facilitated by the multiple APs.”, Tan, paragraphs [0096]–[0098] and [0151]–[0152], Figures 18–19, discloses a quality-of-experience objective based on actual feedback and radio quality states correlated with user-experience indicators. This supplies the claimed quality-of-user-experience score for the managed wireless network.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Tan in view of Siomina.
Claim 8 depends from claim 1.
“The method of claim 1.”
With respect to “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the configuration parameters include at least one or more of a channel selection parameter, a channel width parameter, or a signal power parameter.”, Tan, paragraphs [0063] and [0090], discloses transmit power as a configuration parameter, satisfying the signal-power alternative. The claim’s “at least one or more” and “or” wording does not require channel selection and channel width in addition to signal power.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Tan in view of Siomina.
Claim 9 depends from claim 1.
With respect to “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, Tan, paragraphs [0160], [0164], and [0171], Figure 20, discloses random cell choice and guided-random direction choice among tilt and power changes. Selecting among nonzero tilt-only and power-only directions chooses the identity of a parameter from the selected AP’s collection. This implementation and exclusion of unchanged proposals are explained below.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “The method of claim 1, further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to select Tan’s nonzero, single-parameter directions and discard an unchanged proposal before a live trial. This avoids spending an observation interval on an unchanged AP while retaining stochastic selection among parameter identities and positive or negative changes. A direction’s nonzero coordinate identifies the parameter and its sign determines the change; the claims do not require independent random draws. See Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090].
Regarding claim 10 Tan discloses “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, inherited from claim 9, Tan, paragraphs [0160], [0164], and [0171], Figure 20, discloses random cell choice and guided-random direction choice among tilt and power changes. Selecting among nonzero tilt-only and power-only directions chooses the identity of a parameter from the selected AP’s collection. This implementation and exclusion of unchanged proposals are explained below.
With respect to “wherein making the random change to the configuration of the selected AP by adjusting the configuration parameter to transition from the initial state (S0) to the new state (S1) comprises”, Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090], discloses a guided-random direction and a step along that direction. Its nonzero coordinate chooses the AP parameter and its sign determines the random change producing S1. Selecting nonzero single-parameter moves is the implementation explained below; no separate independent random draw is required.
With respect to “making a random change to the randomly selected configuration parameter of the selected AP to transition from the initial state (S0) to a new state (S1).”, Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090], discloses a guided-random direction and a step along that direction. Its nonzero coordinate chooses the AP parameter and its sign determines the random change producing S1. Selecting nonzero single-parameter moves is the implementation explained below; no separate independent random draw is required.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “The method of claim 1, further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, inherited from claim 9, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “wherein making the random change to the configuration of the selected AP by adjusting the configuration parameter to transition from the initial state (S0) to the new state (S1) comprises”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “making a random change to the randomly selected configuration parameter of the selected AP to transition from the initial state (S0) to a new state (S1).”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to select Tan’s nonzero, single-parameter directions and discard an unchanged proposal before a live trial. This avoids spending an observation interval on an unchanged AP while retaining stochastic selection among parameter identities and positive or negative changes. A direction’s nonzero coordinate identifies the parameter and its sign determines the change; the claims do not require independent random draws. See Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090].
Regarding claim 11, Tan discloses “A centralized coordinator computing system, comprising”, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
Regarding claim 12, Tan discloses “The centralized coordinator computing system of claim 11.”With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is configured to evaluate configuration changes by”, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
Regarding claim 17, Tan discloses “The centralized coordinator computing system of claim 11.”With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the quality score represents the quality of user experience associated with the wireless network facilitated by the multiple APs.”, Tan, paragraphs [0096]–[0098] and [0151]–[0152], Figures 18–19, discloses a quality-of-experience objective based on actual feedback and radio quality states correlated with user-experience indicators. This supplies the claimed quality-of-user-experience score for the managed wireless network.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
Regarding claim 18, Tan discloses “The centralized coordinator computing system of claim 11.”With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the configuration parameters include at least one or more of a channel selection parameter, a channel width parameter, or a signal power parameter.”, Tan, paragraphs [0063] and [0090], discloses transmit power as a configuration parameter, satisfying the signal-power alternative. The claim’s “at least one or more” and “or” wording does not require channel selection and channel width in addition to signal power.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
Regarding claim 19, Tan discloses “The centralized coordinator computing system of claim 11.”With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is further configured to randomly select the configuration parameter from a collection of configuration parameters associated with the selected AP.”, Tan, paragraphs [0160], [0164], and [0171], Figure 20, discloses random cell choice and guided-random direction choice among tilt and power changes. Selecting among nonzero tilt-only and power-only directions chooses the identity of a parameter from the selected AP’s collection. This implementation and exclusion of unchanged proposals are explained below.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “wherein the processor is further configured to randomly select the configuration parameter from a collection of configuration parameters associated with the selected AP.”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
It would have been obvious to select Tan’s nonzero, single-parameter directions and discard an unchanged proposal before a live trial. This avoids spending an observation interval on an unchanged AP while retaining stochastic selection among parameter identities and positive or negative changes. A direction’s nonzero coordinate identifies the parameter and its sign determines the change; the claims do not require independent random draws. See Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090].
Regarding claim 20 Tan discloses “The centralized coordinator computing system of claim 19.”With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is further configured to randomly select the configuration parameter from a collection of configuration parameters associated with the selected AP.”, inherited from claim 19, Tan, paragraphs [0160], [0164], and [0171], Figure 20, discloses random cell choice and guided-random direction choice among tilt and power changes. Selecting among nonzero tilt-only and power-only directions chooses the identity of a parameter from the selected AP’s collection. This implementation and exclusion of unchanged proposals are explained below.
With respect to “wherein the processor is configured to make the random change to the configuration of the selected AP by making the random change to the randomly selected configuration parameter of the selected AP to transition from the initial state (S0) to a new state (S1).”, Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090], discloses a guided-random direction and a step along that direction. Its nonzero coordinate chooses the AP parameter and its sign determines the random change producing S1. Selecting nonzero single-parameter moves is the implementation explained below; no separate independent random draw is required.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “wherein the processor is further configured to randomly select the configuration parameter from a collection of configuration parameters associated with the selected AP.”, inherited from claim 19, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “wherein the processor is configured to make the random change to the configuration of the selected AP by making the random change to the randomly selected configuration parameter of the selected AP to transition from the initial state (S0) to a new state (S1).”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
It would have been obvious to select Tan’s nonzero, single-parameter directions and discard an unchanged proposal before a live trial. This avoids spending an observation interval on an unchanged AP while retaining stochastic selection among parameter identities and positive or negative changes. A direction’s nonzero coordinate identifies the parameter and its sign determines the change; the claims do not require independent random draws. See Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090].
Regarding claim 21, Tan discloses a non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
Regarding claim 22 Tan discloses “The non-transitory computer-readable storage medium of claim 21.”With respect to “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising evaluating configuration changes by”, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
Regarding claim 23, Tan discloses a non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising evaluating configuration changes by”, inherited from claim 22, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 22, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 22, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising applying probabilistic acceptance for suboptimal configurations by”, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 24, Tan discloses “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising evaluating configuration changes by”, inherited from claim 22, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 22, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 22, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising applying probabilistic acceptance for suboptimal configurations by”, inherited from claim 23, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 23, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 23, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 23, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 23, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 23, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 25, Tan discloses “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising evaluating configuration changes by”, inherited from claim 22, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 22, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 22, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising applying probabilistic acceptance for suboptimal configurations by”, inherited from claim 23, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 23, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 24, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 23, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 23, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 23, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 23, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 24, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 27, Tan discloses “The non-transitory computer-readable storage medium of claim 21.”
With respect to “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that the quality score represents the quality of user experience associated with the wireless network facilitated by the multiple APs.”, Tan, paragraphs [0096]–[0098] and [0151]–[0152], Figures 18–19, discloses a quality-of-experience objective based on actual feedback and radio quality states correlated with user-experience indicators. This supplies the claimed quality-of-user-experience score for the managed wireless network.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
Regarding claim 28 , Tan discloses “The non-transitory computer-readable storage medium of claim 21.With respect to “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that the configuration parameters include at least one or more of a channel selection parameter, a channel width parameter, or a signal power parameter.”, Tan, paragraphs [0063] and [0090], discloses transmit power as a configuration parameter, satisfying the signal-power alternative. The claim’s “at least one or more” and “or” wording does not require channel selection and channel width in addition to signal power.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
Regarding claim 29, Tan discloses “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The non-transitory computer-readable storage medium of claim 21 wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, Tan, paragraphs [0160], [0164], and [0171], Figure 20, discloses random cell choice and guided-random direction choice among tilt and power changes. Selecting among nonzero tilt-only and power-only directions chooses the identity of a parameter from the selected AP’s collection. This implementation and exclusion of unchanged proposals are explained below.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “The non-transitory computer-readable storage medium of claim 21 wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
It would have been obvious to select Tan’s nonzero, single-parameter directions and discard an unchanged proposal before a live trial. This avoids spending an observation interval on an unchanged AP while retaining stochastic selection among parameter identities and positive or negative changes. A direction’s nonzero coordinate identifies the parameter and its sign determines the change; the claims do not require independent random draws. See Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090].
Regarding claim 30, Tan discloses “The non-transitory computer-readable storage medium of claim 29.” With respect to “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The non-transitory computer-readable storage medium of claim 21 wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, inherited from claim 29, Tan, paragraphs [0160], [0164], and [0171], Figure 20, discloses random cell choice and guided-random direction choice among tilt and power changes. Selecting among nonzero tilt-only and power-only directions chooses the identity of a parameter from the selected AP’s collection. This implementation and exclusion of unchanged proposals are explained below.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that making the random change to the configuration of the selected AP by adjusting the configuration parameter to transition from the initial state (S0) to the new state (S1) comprises”, Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090], discloses a guided-random direction and a step along that direction. Its nonzero coordinate chooses the AP parameter and its sign determines the random change producing S1. Selecting nonzero single-parameter moves is the implementation explained below; no separate independent random draw is required.
With respect to “making a random change to the randomly selected configuration parameter of the selected AP to transition from the initial state (S0) to a new state (S1).”, Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090], discloses a guided-random direction and a step along that direction. Its nonzero coordinate chooses the AP parameter and its sign determines the random change producing S1. Selecting nonzero single-parameter moves is the implementation explained below; no separate independent random draw is required.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “The non-transitory computer-readable storage medium of claim 21 wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising randomly selecting the configuration parameter from a collection of configuration parameters associated with the selected AP.”, inherited from claim 29, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that making the random change to the configuration of the selected AP by adjusting the configuration parameter to transition from the initial state (S0) to the new state (S1) comprises”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “making a random change to the randomly selected configuration parameter of the selected AP to transition from the initial state (S0) to a new state (S1).”, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
It would have been obvious to select Tan’s nonzero, single-parameter directions and discard an unchanged proposal before a live trial. This avoids spending an observation interval on an unchanged AP while retaining stochastic selection among parameter identities and positive or negative changes. A direction’s nonzero coordinate identifies the parameter and its sign determines the change; the claims do not require independent random draws. See Tan, paragraphs [0164]–[0165], Figure 20, and paragraph [0090].
Claims 3-5, 13-15, and 23-25 are rejected under 35 U.S.C. 103 as being unpatentable over Tan in view of Siomina and further in view of Kirkpatrick (Optimiaztion by Simulated Annealing Science, Vol.220,hereinafter Kirkpatrick).
Regarding claim 3 With respect to “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising evaluating configuration changes by”, inherited from claim 2, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 2, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 2, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “The method of claim 2, further comprising applying probabilistic acceptance for suboptimal configurations by”, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 4, Tan discloses “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising evaluating configuration changes by”, inherited from claim 2, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 2, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 2, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “The method of claim 2, further comprising applying probabilistic acceptance for suboptimal configurations by”, inherited from claim 3, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 3, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 3, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 3, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 3, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 3, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 5, With respect to “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising evaluating configuration changes by”, inherited from claim 2, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 2, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 2, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “The method of claim 2, further comprising applying probabilistic acceptance for suboptimal configurations by”, inherited from claim 3, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 3, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 4, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “The method of claim 4, further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, and adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 3, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 3, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 3, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 3, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 4, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “The method of claim 4, further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, and adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 13, Tan discloses “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is configured to evaluate configuration changes by”, inherited from claim 12, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 12, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 12, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the processor is further configured to apply probabilistic acceptance for suboptimal configurations by”, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 14 ,With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is configured to evaluate configuration changes by”, inherited from claim 12, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 12, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 12, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the processor is further configured to apply probabilistic acceptance for suboptimal configurations by”, inherited from claim 13, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 13, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein the processor is configured to determine the probability of acceptance value for the new state (S1) by setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 13, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 13, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 13, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 13, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein the processor is configured to determine the probability of acceptance value for the new state (S1) by setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Regarding claim 15 , Tan discloses “The centralized coordinator computing system of claim 14.” With respect to “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is configured to evaluate configuration changes by”, inherited from claim 12, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 12, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 12, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the processor is further configured to apply probabilistic acceptance for suboptimal configurations by”, inherited from claim 13, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 13, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein the processor is configured to determine the probability of acceptance value for the new state (S1) by setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 14, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the processor is further configured to repeatedly adjust the AP configurations, evaluate the configuration changes, apply the probabilistic acceptance for the suboptimal configurations, and adjust the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 13, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 13, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 13, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 13, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein the processor is configured to determine the probability of acceptance value for the new state (S1) by setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 14, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the processor is further configured to repeatedly adjust the AP configurations, evaluate the configuration changes, apply the probabilistic acceptance for the suboptimal configurations, and adjust the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
Claims 6,16, and 26 are rejected under 35 U.S.C. 103 as being unpatentable over Tan in view of Siomina and further in view of Kirkpatrick, and further in view of The MathWorks, Inc. (MathWorks), Global Optimization Toolbox User’s Guide, R2021b, September 2021 online revision, Version 4.6, hereinafter MathWorks).
Regarding claim 6, With respect to “A method for configuring multiple wireless access points (APs) in a dense network computing environment, the method comprising”, inherited from claim 1, Tan, paragraphs [0061]–[0063], Figures 1–2, discloses neighboring wireless APs, including Wi-Fi APs, with interacting coverage and power conditions. This meets the stated construction of the dense network environment.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 1, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 1, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 1, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 1, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “The method of claim 1, further comprising evaluating configuration changes by”, inherited from claim 2, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 2, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 2, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “The method of claim 2, further comprising applying probabilistic acceptance for suboptimal configurations by”, inherited from claim 3, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 3, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 4, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “The method of claim 4, further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, and adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, inherited from claim 5, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 1, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 3, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 3, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “The method of claim 5, further comprising determining whether the average change in quality score per iteration is approaching zero.”, as recited. Tan’s cited convergence condition does not expressly identify this averaged-change stopping test.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 1, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 1, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 3, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 3, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 4, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “The method of claim 4, further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, and adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, inherited from claim 5, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
With respect to “The method of claim 5, further comprising determining whether the average change in quality score per iteration is approaching zero.”, MathWorks, Global Optimization Toolbox User’s Guide R2021b, printed page 12-14, PDF page 566, step 5 and “Stopping Conditions for the Algorithm,” tests average objective change over a specified number of iterations against a small function tolerance, default 0.000001. Under Q = −C, the near-zero change test supplies the stated approaching-zero construction.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
It would have been obvious to apply MathWorks’ average-change stopping test to the normalized objective values already produced by the Tan–Siomina–Kirkpatrick loop. Tan already evaluates convergence, and the specified test identifies negligible improvement so that further AP changes and observation intervals can stop after the objective has substantially stabilized. Averaging reduces dependence on one iteration. Reversing the objective sign preserves change magnitude. Success is reasonably expected because the test consumes existing objective values without a new measurement mechanism. Only the stopping test is adopted; MathWorks’ different default acceptance rule is not substituted. See Tan, paragraph [0112]; MathWorks, R2021b, printed page 12-14, PDF page 566.
.
Regarding claim 16, Tan discloses “A centralized coordinator computing system, comprising”, inherited from claim 11, Tan, paragraphs [0246]–[0247], Figure 35, discloses controller 3500, its local-solution generator, and interfaces coordinating the controlled APs.
With respect to “a processor configured to”, inherited from claim 11, Tan, paragraphs [0251]–[0252], Figure 36, discloses processor 3604 executing stored instructions. Configuring it to execute the mapped combined operations is the modification explained in the motivation section for this claim.
With respect to “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measure an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 11, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjust AP configurations by”, inherited from claim 11, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 11, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 11, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the processor is configured to evaluate configuration changes by”, inherited from claim 12, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 12, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 12, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the processor is further configured to apply probabilistic acceptance for suboptimal configurations by”, inherited from claim 13, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 13, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein the processor is configured to determine the probability of acceptance value for the new state (S1) by setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 14, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the processor is further configured to repeatedly adjust the AP configurations, evaluate the configuration changes, apply the probabilistic acceptance for the suboptimal configurations, and adjust the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, inherited from claim 15, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Tan does not expressly disclose the complete limitation “set each of multiple wireless access points (APs) in a dense network computing environment to an initial state (S0)”, inherited from claim 11, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 13, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 13, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “wherein the processor is further configured to determine whether the average change in quality score per iteration is approaching zero.”, as recited. Tan’s cited convergence condition does not expressly identify this averaged-change stopping test.
With respect to “define and adjust the measurement period to balance speed and accuracy of the results”, inherited from claim 11, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 11, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 13, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 13, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein the processor is configured to determine the probability of acceptance value for the new state (S1) by setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 14, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the processor is further configured to repeatedly adjust the AP configurations, evaluate the configuration changes, apply the probabilistic acceptance for the suboptimal configurations, and adjust the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, inherited from claim 15, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
With respect to “wherein the processor is further configured to determine whether the average change in quality score per iteration is approaching zero.”, MathWorks, Global Optimization Toolbox User’s Guide R2021b, printed page 12-14, PDF page 566, step 5 and “Stopping Conditions for the Algorithm,” tests average objective change over a specified number of iterations against a small function tolerance, default 0.000001. Under Q = −C, the near-zero change test supplies the stated approaching-zero construction.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to configure Tan’s coordinator processor 3604 to execute the combined operations, using the controller interfaces and stored instruction execution disclosed in paragraphs [0246]–[0247] and [0251]–[0252]. This implements the duration-controlled optimization through the processor already responsible for the AP configuration loop.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
It would have been obvious to apply MathWorks’ average-change stopping test to the normalized objective values already produced by the Tan–Siomina–Kirkpatrick loop. Tan already evaluates convergence, and the specified test identifies negligible improvement so that further AP changes and observation intervals can stop after the objective has substantially stabilized. Averaging reduces dependence on one iteration. Reversing the objective sign preserves change magnitude. Success is reasonably expected because the test consumes existing objective values without a new measurement mechanism. Only the stopping test is adopted; MathWorks’ different default acceptance rule is not substituted. See Tan, paragraph [0112]; MathWorks, R2021b, printed page 12-14, PDF page 566.
Regarding claim 26, Tan discloses “The non-transitory computer-readable storage medium of claim 25.” With respect to “A non-transitory computer-readable storage medium having stored thereon processor-executable software instructions configured to cause one or more processors to perform operations for configuring multiple wireless access points (APs) in a dense network computing environment, the operations comprising”, inherited from claim 21, Tan, paragraphs [0251]–[0252], Figure 36, discloses memory 3606 storing processor-executable instructions, including non-transitory storage, and processor 3604. Paragraphs [0061]–[0063] and Figures 1–2 supply neighboring interacting wireless APs. The combined operations are implemented by those stored instructions as explained below.
With respect to “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, Tan, paragraphs [0064]–[0067], Figure 3, and paragraphs [0246]–[0247], discloses predefined local settings and configuration delivery to APs. Applying those baseline settings to every managed AP before evaluation is an obvious implementation, not an assertion that identifying an operating solution expressly resets every AP. For the dense environment, see paragraphs [0061]–[0063].
With respect to “measuring an initial quality score (Q0) for the initial state (S0) over a measurement period”, inherited from claim 21, Tan, paragraphs [0064] and [0088]–[0089], evaluates the initial solution using live reports collected during a test period and normalizes cost by report count. Define Q0 = −C0 so increased quality corresponds to decreased cost, preserving the measured information and ranking.
With respect to “iteratively adjusting AP configurations by”, inherited from claim 21, Tan, paragraph [0112], Table I, and Figure 6, repeats candidate generation, configuration, evaluation, and selection of the next reference state through its optimization loop.
With respect to “randomly selecting an AP from the multiple APs”, inherited from claim 21, Tan, paragraphs [0061], [0090], and [0160], discloses random selection of the cell to adjust and a fixed selected-cell count. Selecting one cell in the disclosed AP implementation supplies random selection of an AP from the managed APs.
With respect to “making a random change to the configuration of the selected AP by adjusting a configuration parameter to transition from the initial state (S0) to a new state (S1)”, inherited from claim 21, Tan, paragraphs [0066]–[0067] and [0090], discloses perturbing a selected node’s configuration parameter, including random direction and step choices, and applying the candidate settings. The current and candidate local configurations supply S0 and S1.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising evaluating configuration changes by”, inherited from claim 22, Tan, Table I and paragraph [0112], evaluates configuration changes through current-versus-candidate objective comparison and acceptance. For processor execution and stored instructions, see paragraphs [0251]–[0252].
With respect to “comparing the new quality score (Q1) with the initial quality score (Q0)”, inherited from claim 22, Tan, Table I and paragraph [0112], compares candidate cost C1 with current cost C0. Under Q = −C, this compares Q1 and Q0 with reversed improvement direction.
With respect to “accepting the new state (S1) as the new basis for subsequent configuration adjustments in response to determining that the new quality score (Q1) is greater than the initial quality score (Q0).”, inherited from claim 22, Tan, Table I and paragraph [0112], accepts a lower-cost candidate and uses it as the next reference solution. Under Q = −C, C1 < C0 is Q1 > Q0, so S1 becomes the basis for subsequent configuration adjustments.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising applying probabilistic acceptance for suboptimal configurations by”, inherited from claim 23, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “determining a probability of acceptance value for the new state (S1) in response to determining that the new quality score (Q1) is not greater than the initial quality score (Q0)”, inherited from claim 23, Tan, Table I and paragraph [0112], provides probabilistic acceptance of non-improving candidate configurations using its annealing acceptance rule. With Q = −C, a candidate whose Q1 is not greater than Q0 corresponds to a non-improving cost, including equality. Processor execution is disclosed in paragraphs [0251]–[0252].
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 24, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, inherited from claim 25, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
Tan does not expressly disclose the complete limitation “setting each of the multiple APs to an initial state (S0)”, inherited from claim 21, as recited. The complete implementation stated in this limitation is treated as a reasoned modification of Tan’s disclosed settings or candidate-generation options, rather than an express disclosure of the complete claimed operation.
Tan does not expressly disclose the complete limitation “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, as recited. The cited Tan disclosure does not expressly provide the claimed duration adjustment for the speed and accuracy trade-off.
Tan does not expressly disclose the complete limitation “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, as recited. Tan supplies post-change evaluation but does not expressly tie it to the claimed defined and adjusted observation period.
Tan does not expressly disclose the complete limitation “generating a random number between 0 and 1”, inherited from claim 23, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 23, as recited. The cited Tan acceptance-probability disclosure does not expressly identify this uniform-random-number implementation.
Tan does not expressly disclose the complete limitation “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising determining whether the average change in quality score per iteration is approaching zero.”, as recited. Tan’s cited convergence condition does not expressly identify this averaged-change stopping test.
With respect to “defining and adjusting the measurement period to balance speed and accuracy of the results”, inherited from claim 21, Siomina, Figure 3, steps 302–310, discloses setting and adjusting measurement requirements and measuring accordingly. Figure 6, steps 602–606, and claims 10–11 disclose adjustment of observation duration for measurement capacity and accuracy. In Tan’s evaluation stage, this balances observation delay against measurement reliability.
With respect to “measuring a new quality score (Q1) for the new state (S1) using the defined and adjusted measurement period.”, inherited from claim 21, Tan, paragraphs [0067], [0091], and [0112], discloses post-change measurement reports and candidate evaluation. Siomina, Figure 3, step 310, and Figure 6, steps 602–606, supplies measurements under adjusted-duration requirements. The combination collects candidate-state observations during that interval and computes Q1 = −C1 using the same normalized objective.
With respect to “generating a random number between 0 and 1”, inherited from claim 23, Kirkpatrick, printed page 672, discloses generating a uniform random number in (0,1) for the Metropolis acceptance decision. Tan’s cited probability rule alone is not relied upon as express disclosure of this comparator implementation.
With respect to “accepting the new state (S1) based on whether the random number is less than the determined probability of acceptance value.”, inherited from claim 23, Kirkpatrick, printed page 672, accepts the proposed configuration when the uniform random number is below the computed acceptance probability. Applied to Tan’s candidate AP configuration, this accepts S1; the probability is one at equal scores.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations such that determining the probability of acceptance value for the new state (S1) comprises setting the probability of acceptance value equal to e(Q1-Q0)/T in which T is a tolerance parameter that is indicative of the system tolerance for accepting suboptimal states.”, inherited from claim 24, Tan, Table I and paragraph [0112], supplies exponential acceptance using the cost difference and temperature T; Kirkpatrick, printed page 672, corroborates the rule. With Q = −C, C0 − C1 = Q1 − Q0, giving exp((Q1 − Q0)/T). T controls tolerance for non-improving states. This exponential construction follows the specification’s description of Figure 2, block 212, and the accompanying numerical example.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising repeatedly adjusting the AP configurations, evaluating the configuration changes, applying the probabilistic acceptance for the suboptimal configurations, adjusting the tolerance parameter (T) over time to reduce the system tolerance for suboptimal states as the system approaches a threshold value indicating optimal configuration.”, inherited from claim 25, Tan, paragraphs [0107] and [0112], Table I, and Figure 6, repeats AP adjustment, evaluation, and probabilistic acceptance while updating temperature toward convergence. Kirkpatrick, printed pages 672–673, explains gradual cooling. Reducing T decreases acceptance of lower-quality states as the convergence threshold is approached; no guarantee of global optimality is asserted.
With respect to “wherein the stored processor-executable software instructions are configured to cause a processor to perform operations further comprising determining whether the average change in quality score per iteration is approaching zero.”, MathWorks, Global Optimization Toolbox User’s Guide R2021b, printed page 12-14, PDF page 566, step 5 and “Stopping Conditions for the Algorithm,” tests average objective change over a specified number of iterations against a small function tolerance, default 0.000001. Under Q = −C, the near-zero change test supplies the stated approaching-zero construction.
Before the effective filing date, it would have been obvious to apply Tan’s predefined local settings to every managed AP through its disclosed controller interface before collecting baseline reports. This establishes a measured reference state corresponding to the intended starting configuration and predictably uses the same interface used for later trials. See Tan, paragraphs [0064]–[0067] and [0246]–[0247].
It would have been obvious to apply Siomina’s measurement-duration adjustment to the observation interval feeding Tan’s live evaluation stage. Longer observation delays the next configuration decision; shorter observation can reduce measurement reliability. Siomina’s duration and accuracy control addresses that conflict in radio measurements serving network management. The controller defines an initial interval for Q0, adjusts the observation duration according to measurement capability and accuracy considerations, and uses that interval for Q1. See Tan, paragraphs [0064], [0067], [0088]–[0091]; Siomina, Figure 3, steps 302–310, Figure 6, steps 602–606, and claims 10–11.
A person of ordinary skill would reasonably expect success because the modification controls the duration of observations already used by Tan’s evaluator. Retaining the same report-normalized objective C, weights, and thresholds and defining Q = −C preserves the comparison across trials. Normalization prevents additional report count alone from increasing the score; it does not eliminate traffic variation or sampling error. The expected result is a controllable observation-delay and accuracy trade-off with the existing candidate-generation and comparison functions. See KSR International Co. v. Teleflex Inc., 550 U.S. 398, 417–418 (2007), and MPEP §2143.
It would have been obvious to store instructions implementing the combined operations in Tan’s non-transitory memory 3606 for execution by processor 3604. Tan already performs the AP coordination through stored instructions, paragraphs [0251]–[0252], so storing the added duration-control and mapped decision operations uses that disclosed execution arrangement predictably.
It would have been obvious to implement Tan’s probabilistic acceptance decision with Kirkpatrick’s uniform-random-number comparison. Kirkpatrick is pertinent to implementing the simulated-annealing procedure used by Tan. A draw in (0,1) accepted when below the exponential probability realizes that probability without changing the measured objective or candidate generation. When Q1 equals Q0, the exponential probability is one and the candidate is accepted. See Tan, Table I and paragraph [0112]; Kirkpatrick, printed page 672.
It would have been obvious to apply MathWorks’ average-change stopping test to the normalized objective values already produced by the Tan–Siomina–Kirkpatrick loop. Tan already evaluates convergence, and the specified test identifies negligible improvement so that further AP changes and observation intervals can stop after the objective has substantially stabilized. Averaging reduces dependence on one iteration. Reversing the objective sign preserves change magnitude. Success is reasonably expected because the test consumes existing objective values without a new measurement mechanism. Only the stopping test is adopted; MathWorks’ different default acceptance rule is not substituted. See Tan, paragraph [0112]; MathWorks, R2021b, printed page 12-14, PDF page 566.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANGEL T BROCKMAN whose telephone number is (571)270-5664. The examiner can normally be reached Monday-Thursday 6:00 AM-4:30 PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Charles Jiang can be reached at 571-270-7191. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ANGEL T BROCKMAN/Examiner, Art Unit 2412
Prosecution Insights