Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Detailed Action
2. This office action is in response to the filing with the office dated 10/29/2025.
Response to Applicant’s arguments
3. Applicant’s arguments and claim amendments filed with the office on 10/29/2025 have been fully considered and found to be non-persuasive. Regarding applicant’s arguments about test and measurement port of Hazzard et al (US 2022/0390513 A1) please see, ([0025) although illustrated in FIG. 2 as only having a single output, the multiplexer 232 may include multiple outputs, especially when there are multiple inputs to the multiplexer from the DUT 110, such as eight, sixteen, thirty-two, or more inputs. Increasing the number of outputs of the multiplexer 232 has the effect of increasing the bandwidth, and therefore the testing speed, of the test and measurement system 200). Also see the 35 U.S.C. 112 rejection regarding the claim limitation recitation” Partly connecting”.
Applicant's amendment necessitated the ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Claim Rejections – 35 U.S.C. 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
4. Claim 1, 11 and 21 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claims 1, 11 and 21 recite the limitation, wherein the plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit.
It is not clear It is not clear to one of the ordinary skill in the art, what is meant by partly connecting the plurality of connectors of the device under test with the front end ports of the switch unit. It is well known in the art to enable/disable connection through a switch or software. The specification does not provide any details about how “plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit“. Appropriate correction to the claim language is required to clarify this limitation.
Claims 3-5, 7-10, 13, 14, 16-20 are rejected under 35 U.S.C. 112 (a) due to their dependency on the independent claim.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
5. Claim 1 and 11 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 and 11 recites the limitation, wherein the plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit. It is not clear to one of the ordinary skill in the art, what is meant by partly connecting the plurality of connectors of the device under test with the front end ports of the switch unit, which renders the claim limitation indefinite. It is well known in the art to enable/disable connection through a switch or software. In the absence of any details in the specification about how “plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit“, it is difficult for one of the ordinary skill in the art to understand how the artificial intelligence based cable optimization method is implemented through partly connected cable connectors. Appropriate correction to the claim language is required to clarify this limitation.
Claims 3-5, 7-10, 13, 14, 16-20 are rejected under 35 U.S.C. 112 (b) due to their dependency on the independent claim.
Claim Rejections – 35 U.S.C. 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
PNG
media_image1.png
725
469
media_image1.png
Greyscale
6. Claims 1, 3-5, 7-11, 13, 14, 16-21 are rejected under 35 U.S.C. 103 as being unpatentable over Hazzard et al (US 2022/0390513 A1) and in view of Ren et al (US 2021/0342516 A1).
Regarding independent claim 1, Hazzard et al (US 2022/0390513 A1) teaches, A method of optimizing cabling of at least one device under test (paragraph [0002]) when performing tests on the at least one device under test, wherein the method comprises the steps of: providing at least one device under test (element 110, figure 1, paragraph [001]) with a plurality of connectors (elements 121, 123, 125, 127, 141, 143, 145 and 147, figure 1, paragraph [0011]), wherein the connectors are capable of supporting different frequencies and/or different signal directions ([0015] Depending on the DUT 110 and the test and measurement instrument 160, signals may travel both to and from each of the DUT 110 and the test and measurement instrument 160), providing at least one testing device with a plurality of test ports (element 162 of the testing instrument 160 as shown in figure 1, paragraphs [0025], [0061]) , wherein the test ports are capable of supporting
PNG
media_image2.png
480
719
media_image2.png
Greyscale
different frequencies and/or different signal directions (paragraphs [0004], [0015]), wherein the at least one device under test and the testing device are to be connected with each other via a cabling (RF cable 152, figure 1, paragraph [0014], providing a switch unit that is interposed between the testing device and the device under test, wherein the switch unit has front end ports as well as back end ports (the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130. (paragraph [0017]), wherein the plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit, whereas the back end ports of the switch unit are connected with the plurality of test ports of the testing device (figures 1 and 2, paragraphs [0019], [0025]), wherein the cabling comprises connecting cables between the plurality of connectors and the front end ports as well as connecting cables between the plurality of test ports and the back end ports (connectors 121, 123, 125 and 127, 141, 143, 145, 147, and cables 131, 133, 135 and 137 ae shown in figure 1, paragraphs [0019], [0020]; although illustrated in FIG. 2 as only having a single output, the multiplexer 232 may include multiple outputs, especially when there are multiple inputs to the multiplexer from the DUT 110, such as eight, sixteen, thirty-two, or more inputs. Increasing the number of outputs of the multiplexer 232 has the effect of increasing the bandwidth, and therefore the testing speed, of the test and measurement system 200 [0025]) and wherein the processing circuit sends control signals to the switch unit to adapt an internal signal routing, thereby affecting the cabling in order to obtain the optimized cabling during the tests performed (paragraph [0019]).
Hazzard et al further teaches, ([0017] the user may program the measurement instrument 160 to cause a first signal from the DUT 110, for example Lane 0, signal A, to be selected for testing, and then test desired parameters of the selected signal at the measurement instrument 160 for a period of time. After the first testing period, the testing program causes the multiplexer 132 of the input selector 130 to select a second signal from the DUT 110 for testing, for example Lane 0, signal B. In this manner, all of the signals from the DUT 110 may be scripted to be tested without any necessity of the user to physically change any cables between the DUT 110 and the measurement instrument 160. Instead, the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130).
Hazzard et al does not explicitly teach optimizing the cabling between the at least one device under test and the at least one testing device by a processing circuit.
Ren et al (US 2021/0342516 A1) teaches, (a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract). The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections (paragraph [0026]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 3, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 1.
Hazzard et al (US 2022/0390513 A1) further teaches, (input selector 130 is embodied as an Application Specific Integrated Circuit (ASIC) or a Micro-Electro Mechanical System (MEMS) switch. In other embodiments the input selector 130 may include multiple discrete components mounted to a Printed Circuit (PC) board (paragraph [0016]) [0038]).
Hazzard fails to explicitly teach, wherein the processing circuit executes an algorithm which, when executed on the processing circuit, optimizes the cabling between the at least one device under test and the testing device.
Ren et al (US 2021/0342516 A1) further teaches, wherein the processing circuit executes an algorithm which, when executed on the processing circuit, optimizes the cabling between the at least one device under test and the testing device ([0025] FIG. 1 illustrates a block diagram of an example of a circuit routing system 100 constructed according to the principles of the disclosure. The circuit routing system 100 is a RL system that determines optimal connection routes between terminal groups of a circuit's nets. The circuit routing system 100 can determine connection routes for all of a circuit's nets or less than all of the circuit's nets. For example, other types of routing algorithms or systems can be used for routing one or more of the nets and the circuit routing system 100 can be used for the remaining nets. The circuit routing system 100 includes a routing environment 110 and a RL agent 120. The circuit routing system 100 is configured to model the routing process for a circuit as a game process for the RL agent 120 using the routing environment 110. Also see paragraphs [0026], [0030], [0031], [0032]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 4, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 3.
Hazzard et al (US 2022/0390513 A1) fails to teach, wherein the algorithm is an artificial intelligence algorithm.
Ren et al (US 2021/0342516 A1) further teaches, wherein the algorithm is an artificial intelligence algorithm (figures 1, 2 and paragraphs [0030], [0031], [0092]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements using an algorithm as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, the NNs can be used to determine connections between terminals groups of the nets of circuits, as taught by Ren et al (paragraph [0091], [0092]).
Regarding dependent claim 5, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 4.
Hazzard et al (US 2022/0390513 A1) fails to teach, wherein the artificial intelligence algorithm uses reinforcement learning, supervised learning and/or reward function with optimization.
Ren et al (US 2021/0342516 A1) further teaches, wherein the artificial intelligence algorithm uses reinforcement learning (paragraphs [0020], [0034], [0059]), supervised learning and/or reward function with optimization (paragraphs [0030], figure 3 and its description in paragraphs [0048], [0049]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements using an algorithm as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to use reinforcement learning (RL), such as deep RL, to determine routing of connections for circuits, an agent is trained to complete a task of routing connections for a circuit within a routing environment. The agent sends routing actions to the routing environment and in return receives observations and rewards from the routing environment. The rewards provide a measurement of the success of a routing action with respect to completing a routing task within the routing environment, as taught by Ren et al (paragraph [0020]).
Regarding dependent claim 7, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 1.
Hazzard et al (US 2022/0390513 A1) further teaches, wherein the processing circuit considers the switch unit when optimizing the cabling between the at least one device under test and the at least one testing device ([0019] In some examples, the input selector 130 may include a memory 136 and/or a controller or processor 138. The input selector 130 may be tested during manufacturing or at the factory to measure the effects of including the input selector 130 in the testing system 100 compared to connecting a particular output of the DUT 110 directly to the test port 162 of the measurement instrument 160. Compensating for any negative effects of including the input selector 130, its components including the multiplexer 132 and amplifier 134, and its related cables 131, 133, 135, 137, and 152 is referred to as de-embedding or calibrating. The term calibration parameters in this disclosure refers to any calibration parameter, including de-embed parameters of the input selector 130, that are used to remove effects or impacts of the presence of the input selector 130 and its related cables 131, 133, 135, 137, and 152 in the testing system 100 from any signal measurements made by the test and measurement instrument 160. The calibration parameters may be stored in a particularized calibration parameters memory 140 or in the general memory 136 and sent to the test and measurement instrument 160 during a testing session. In some examples, the calibration parameters may be sent to an analysis device that is remote from the test and measurement instrument 160. The analysis device may collect the calibration parameters as well as data from the test and measurement instrument 160 and provide any processing needed of the data, as will be understood by one skilled in the art. In other examples, the calibration parameters for a particular input selector 130 may be stored in a memory 163 located in the test and measurement instrument 160 or retrieved from remote storage, such as cloud storage and sent to the input selector 130 to be stored in its calibration parameters storage 140. The particular calibration parameters used to de-embed the effects of the input selector 130 may be identified based on a serial number or other identification number of the input selector 130, for example. [0020] Scattering parameters, also referred to as S-parameters, for each of the ports of the multiplexer 132 may be stored along with the calibration parameters 140 or in the separate memory 136 to assist the test and measurement instrument 160 in de-embedding the input selector 130 from the signal of the DUT 110. The through path and isolation path S-parameters are stored for each of the connectors 141, 143, 145, 147 of the multiplexer 132 as well as the port 162 of the measurement instrument 160. In some examples, a user may upload or otherwise port their own S-parameters, such as if the user wishes to change the S-parameters or add stress to the system).
Regarding dependent claim 8, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 1.
Hazzard et al (US 2022/0390513 A1) further teaches, wherein the processing circuit considers the at least one device under test, the at least one testing device when optimizing the cabling between the at least one device under test and the at least one testing device ([0017] all of the signals from the DUT 110 may be scripted to be tested without any necessity of the user to physically change any cables between the DUT 110 and the measurement instrument 160. Instead, the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130; [0019] In some examples, the input selector 130 may include a memory 136 and/or a controller or processor 138. The input selector 130 may be tested during manufacturing or at the factory to measure the effects of including the input selector 130 in the testing system 100 compared to connecting a particular output of the DUT 110 directly to the test port 162 of the measurement instrument 160. Compensating for any negative effects of including the input selector 130, its components including the multiplexer 132 and amplifier 134, and its related cables 131, 133, 135, 137, and 152 is referred to as de-embedding or calibrating. The term calibration parameters in this disclosure refers to any calibration parameter, including de-embed parameters of the input selector 130, that are used to remove effects or impacts of the presence of the input selector 130 and its related cables 131, 133, 135, 137, and 152 in the testing system 100 from any signal measurements made by the test and measurement instrument 160. The calibration parameters may be stored in a particularized calibration parameters memory 140 or in the general memory 136 and sent to the test and measurement instrument 160 during a testing session. In some examples, the calibration parameters may be sent to an analysis device that is remote from the test and measurement instrument 160. The analysis device may collect the calibration parameters as well as data from the test and measurement instrument 160 and provide any processing needed of the data, as will be understood by one skilled in the art. In other examples, the calibration parameters for a particular input selector 130 may be stored in a memory 163 located in the test and measurement instrument 160 or retrieved from remote storage, such as cloud storage and sent to the input selector 130 to be stored in its calibration parameters storage 140. The particular calibration parameters used to de-embed the effects of the input selector 130 may be identified based on a serial number or other identification number of the input selector 130, for example. [0020] Scattering parameters, also referred to as S-parameters, for each of the ports of the multiplexer 132 may be stored along with the calibration parameters 140 or in the separate memory 136 to assist the test and measurement instrument 160 in de-embedding the input selector 130 from the signal of the DUT 110. The through path and isolation path S-parameters are stored for each of the connectors 141, 143, 145, 147 of the multiplexer 132 as well as the port 162 of the measurement instrument 160. In some examples, a user may upload or otherwise port their own S-parameters, such as if the user wishes to change the S-parameters or add stress to the system).
Hazzard et al does not explicitly teach test plan comprising planned tests.
Ren et al (US 2021/0342516 A1) teaches, (a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract).
[0006] In another aspect, the disclosure provides a method of training a RL agent, employing a game, to determine routing connections for circuits. In one example, this method includes: (1) observing, by the RL agent, a current routing state between terminal positions of a circuit in a routing environment, (2) providing, from the RL agent to the routing environment, a routing action that changes the current routing state between the terminal positions, wherein the RL agent provides the routing action based on a model for routing nets that is independent of a number of the nets of the circuit, (3) evaluating, by the routing environment, the routing action, and (4) providing, from the routing environment to the RL agent, one or more reward based on the evaluating.
[0020] In contrast to the limitations of existing methods, the disclosure provides a general solution to determining connections between terminals of circuits. Disclosed herein is a machine learning (ML) based method that uses reinforcement learning (RL), such as deep RL, to determine routing of connections for circuits. The disclosed ML method leverages RL to optimize routing of circuit connections using a game process. In the method, an agent is trained to complete a task of routing connections for a circuit within a routing environment. The agent sends routing actions to the routing environment and in return receives observations and rewards from the routing environment. The rewards provide a measurement of the success of a routing action with respect to completing a routing task within the routing environment).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 9, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 8.
Hazzard et al (US 2022/0390513 A1) further teaches, ([0014] The multiplexer 132 of the input selector 130 provides a selection function to the user. In other words, the user controls the multiplexer 132 of the input selector 130 to choose any of the four inputs to the multiplexer as the desired output of the input selector. The selected signal from the multiplexer 132 is amplified by amplifier 134 before sending it through an RF connection, such as a cable 152, to a test port 162 of the measurement instrument 160. Although in FIG. 1 the amplifier 134 includes only a single output being sent to the measurement instrument 160, in other embodiments the amplifier 134 may send more than one signal to the measurement instrument 160 for testing. For example, the amplifier 134 may be a differential amplifier that sends a differential signal over two separate paths to the measurement instrument 160. In other embodiments the input selector 130 may include multiple amplifiers 134, which each send an amplified signal to the measurement instrument 160. For example the input selector 130 may be controlled to simultaneously send the A input of both Lanes 0 and Lane 1, as separate signals routed from the multiplexer 132 through separate amplifiers 134, to the measurement instrument 160. Or, the input selector 130 may be controlled to simultaneously send both the A and B inputs of Lane 1 out of the multiplexer 132 and individual amplifiers 134, as separate signals, to the measurement instrument 160).
Hazard et al fails to explicitly teach, wherein the processing circuit returns a scheduling plan with an optimized order of the planned tests.
Ren et al (US 2021/0342516 A1) teaches, (a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract). [0006] In another aspect, the disclosure provides a method of training a RL agent, employing a game, to determine routing connections for circuits. In one example, this method includes: (1) observing, by the RL agent, a current routing state between terminal positions of a circuit in a routing environment, (2) providing, from the RL agent to the routing environment, a routing action that changes the current routing state between the terminal positions, wherein the RL agent provides the routing action based on a model for routing nets that is independent of a number of the nets of the circuit, (3) evaluating, by the routing environment, the routing action, and (4) providing, from the routing environment to the RL agent, one or more reward based on the evaluating. [0020] In contrast to the limitations of existing methods, the disclosure provides a general solution to determining connections between terminals of circuits. Disclosed herein is a machine learning (ML) based method that uses reinforcement learning (RL), such as deep RL, to determine routing of connections for circuits. The disclosed ML method leverages RL to optimize routing of circuit connections using a game process. In the method, an agent is trained to complete a task of routing connections for a circuit within a routing environment. The agent sends routing actions to the routing environment and in return receives observations and rewards from the routing environment. The rewards provide a measurement of the success of a routing action with respect to completing a routing task within the routing environment. [0026] The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 10, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the method according to claim 1.
Hazzard et al (US 2022/0390513 A1) further teaches ([0037] one or more test and measurement instruments to test one or more signal lanes which may come from one or more DUTs. The input selectors of the disclosure can allow for switching between the various signal lanes to one or more test and measurement instruments. The input selectors have known calibration parameters determined at manufacturing or the factory that can be used by the test and measurement instruments to remove any effects of the selectors, their components, and their connection cables. Operation of the input selector may be programmatically controlled, and a user may program a selector to step through all of the connected signal lanes coming from a DUT without requiring any cables to be disconnected or rearranged. This can save hours of manual labor during a testing session).
Hazzard et al does not explicitly teach wherein the processing circuit returns a cabling plan that provides the optimized cabling.
Ren et al (US 2021/0342516 A1) further teaches, wherein the processing circuit returns a cabling plan that provides the optimized cabling ([0026] The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections). [0080] The routed nets from the method 700 can be used in a design flow to produce the circuit. For example, the routed nets can be used in the physical design portion of a design flow that includes planning the layout of the different electrical components, placing the electrical components, and determining the routing of the connections between terminals of the electrical components).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
PNG
media_image1.png
725
469
media_image1.png
Greyscale
Regarding independent claim 11, Hazzard et al (US 2022/0390513 A1) teaches, A system for performing tests on at least one device under test (paragraph [0010], wherein the system comprises at least one device under test (element 110, figure 1, paragraph [001]) with a plurality of connectors (elements 121, 123, 125, 127, 141, 143, 145 and 147, figure 1, paragraph [0011]), wherein the connectors are capable of supporting different frequencies and/or different signal directions ([0015] Depending on the DUT 110 and the test and measurement instrument 160, signals may travel both to and from each of the DUT 110 and the test and measurement instrument 160), wherein the system further comprises at least one testing device with a plurality of test ports (element 162 of the testing instrument 160 as shown in figure 1, paragraphs [0025], [0061]), wherein the test ports are capable of supporting different frequencies and/or different signal directions (paragraphs [0004], [0015]), wherein the at least one device under test and the testing device are to be connected with each other via a cabling (RF cable 152, figure 1, paragraph [0014]), wherein a switch unit is provided that is connected with the at least one device under test and the at least one testing device, wherein the switch unit has front end ports as well as back end ports (the switching to connect the desired signal of the DUT 110 to
PNG
media_image2.png
480
719
media_image2.png
Greyscale
the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130. (paragraph [0017]), wherein the plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit, whereas the back end ports of the switch unit are connected with the plurality of test ports of the testing device (figures 1 and 2, paragraphs [0019], [0025]), wherein the cabling comprises connecting cables between the plurality of connectors and the front end ports as well as connecting cables between the plurality of test ports and the back end ports (connectors 121, 123, 125 and 127, 141, 143, 145, 147, and cables 131, 133, 135 and 137 ae shown in figure 1, paragraphs [0019], [0020]; although illustrated in FIG. 2 as only having a single output, the multiplexer 232 may include multiple outputs, especially when there are multiple inputs to the multiplexer from the DUT 110, such as eight, sixteen, thirty-two, or more inputs. Increasing the number of outputs of the multiplexer 232 has the effect of increasing the bandwidth, and therefore the testing speed, of the test and measurement system 200 [0025]) and wherein the processing circuit sends control signals to the switch unit to adapt an internal signal routing, thereby affecting the cabling in order to obtain the optimized cabling during the tests performed (paragraph [0019]).
Hazzard et al further teaches, ([0017] the user may program the measurement instrument 160 to cause a first signal from the DUT 110, for example Lane 0, signal A, to be selected for testing, and then test desired parameters of the selected signal at the measurement instrument 160 for a period of time. After the first testing period, the testing program causes the multiplexer 132 of the input selector 130 to select a second signal from the DUT 110 for testing, for example Lane 0, signal B. In this manner, all of the signals from the DUT 110 may be scripted to be tested without any necessity of the user to physically change any cables between the DUT 110 and the measurement instrument 160. Instead, the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130).
Hazzard et al does not explicitly teach wherein the system comprises a processing circuit that optimizes the cabling between the at least one device under test and the at least one testing device.
Ren et al (US 2021/0342516 A1) teaches, a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract). The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections (paragraph [0026]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 13, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 11.
Hazzard et al (US 2022/0390513 A1) further teaches, (input selector 130 is embodied as an Application Specific Integrated Circuit (ASIC) or a Micro-Electro Mechanical System (MEMS) switch. In other embodiments the input selector 130 may include multiple discrete components mounted to a Printed Circuit (PC) board (paragraph [0016]) [0038]).
Hazzard fails to explicitly teach, wherein the processing circuit is executes an algorithm which, when executed on the processing circuit, optimizes the cabling between the at least one device under test and the testing device.
Ren et al (US 2021/0342516 A1) further teaches, wherein the processing circuit executes an algorithm which, when executed on the processing circuit, optimizes the cabling between the at least one device under test and the testing device ([0025] FIG. 1 illustrates a block diagram of an example of a circuit routing system 100 constructed according to the principles of the disclosure. The circuit routing system 100 is a RL system that determines optimal connection routes between terminal groups of a circuit's nets. The circuit routing system 100 can determine connection routes for all of a circuit's nets or less than all of the circuit's nets. For example, other types of routing algorithms or systems can be used for routing one or more of the nets and the circuit routing system 100 can be used for the remaining nets. The circuit routing system 100 includes a routing environment 110 and a RL agent 120. The circuit routing system 100 is configured to model the routing process for a circuit as a game process for the RL agent 120 using the routing environment 110. Also see paragraphs [0026], [0030], [0031], [0032]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 14, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 13.
Hazzard et al (US 2022/0390513 A1) fails to teach, wherein the algorithm is an artificial intelligence algorithm.
Ren et al (US 2021/0342516 A1) further teaches, wherein the algorithm is an artificial intelligence algorithm (figures 1, 2 and paragraphs [0030], [0031], [0092]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements using an algorithm as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, the NNs can be used to determine connections between terminals groups of the nets of circuits, as taught by Ren et al (paragraph [0091], [0092]).
Regarding dependent claim 16, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 11.
Hazzard et al (US 2022/0390513 A1) further teaches, wherein the switch unit comprises at least one combiner, a signal switching and conditioning unit, an amplifier, a splitter, a switch and/or an attenuator (figure 1, paragraph [0010]).
Regarding dependent claim 17, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 11.
Hazzard et al (US 2022/0390513 A1) further teaches, wherein the processing circuit considers the switch unit when optimizing the cabling between the at least one device under test and the at least one testing device ([0019] In some examples, the input selector 130 may include a memory 136 and/or a controller or processor 138. The input selector 130 may be tested during manufacturing or at the factory to measure the effects of including the input selector 130 in the testing system 100 compared to connecting a particular output of the DUT 110 directly to the test port 162 of the measurement instrument 160. Compensating for any negative effects of including the input selector 130, its components including the multiplexer 132 and amplifier 134, and its related cables 131, 133, 135, 137, and 152 is referred to as de-embedding or calibrating. The term calibration parameters in this disclosure refers to any calibration parameter, including de-embed parameters of the input selector 130, that are used to remove effects or impacts of the presence of the input selector 130 and its related cables 131, 133, 135, 137, and 152 in the testing system 100 from any signal measurements made by the test and measurement instrument 160. The calibration parameters may be stored in a particularized calibration parameters memory 140 or in the general memory 136 and sent to the test and measurement instrument 160 during a testing session. In some examples, the calibration parameters may be sent to an analysis device that is remote from the test and measurement instrument 160. The analysis device may collect the calibration parameters as well as data from the test and measurement instrument 160 and provide any processing needed of the data, as will be understood by one skilled in the art. In other examples, the calibration parameters for a particular input selector 130 may be stored in a memory 163 located in the test and measurement instrument 160 or retrieved from remote storage, such as cloud storage and sent to the input selector 130 to be stored in its calibration parameters storage 140. The particular calibration parameters used to de-embed the effects of the input selector 130 may be identified based on a serial number or other identification number of the input selector 130, for example. [0020] Scattering parameters, also referred to as S-parameters, for each of the ports of the multiplexer 132 may be stored along with the calibration parameters 140 or in the separate memory 136 to assist the test and measurement instrument 160 in de-embedding the input selector 130 from the signal of the DUT 110. The through path and isolation path S-parameters are stored for each of the connectors 141, 143, 145, 147 of the multiplexer 132 as well as the port 162 of the measurement instrument 160. In some examples, a user may upload or otherwise port their own S-parameters, such as if the user wishes to change the S-parameters or add stress to the system).
Regarding dependent claim 18, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 11.
Hazzard et al (US 2022/0390513 A1) further teaches, wherein the processing circuit considers the at least one device under test, the at least one testing device as well as a test plan comprising planned tests when optimizing the cabling between the at least one device under test and the at least one testing device ([0017] all of the signals from the DUT 110 may be scripted to be tested without any necessity of the user to physically change any cables between the DUT 110 and the measurement instrument 160. Instead, the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130; [0019] In some examples, the input selector 130 may include a memory 136 and/or a controller or processor 138. The input selector 130 may be tested during manufacturing or at the factory to measure the effects of including the input selector 130 in the testing system 100 compared to connecting a particular output of the DUT 110 directly to the test port 162 of the measurement instrument 160. Compensating for any negative effects of including the input selector 130, its components including the multiplexer 132 and amplifier 134, and its related cables 131, 133, 135, 137, and 152 is referred to as de-embedding or calibrating. The term calibration parameters in this disclosure refers to any calibration parameter, including de-embed parameters of the input selector 130, that are used to remove effects or impacts of the presence of the input selector 130 and its related cables 131, 133, 135, 137, and 152 in the testing system 100 from any signal measurements made by the test and measurement instrument 160. The calibration parameters may be stored in a particularized calibration parameters memory 140 or in the general memory 136 and sent to the test and measurement instrument 160 during a testing session. In some examples, the calibration parameters may be sent to an analysis device that is remote from the test and measurement instrument 160. The analysis device may collect the calibration parameters as well as data from the test and measurement instrument 160 and provide any processing needed of the data, as will be understood by one skilled in the art. In other examples, the calibration parameters for a particular input selector 130 may be stored in a memory 163 located in the test and measurement instrument 160 or retrieved from remote storage, such as cloud storage and sent to the input selector 130 to be stored in its calibration parameters storage 140. The particular calibration parameters used to de-embed the effects of the input selector 130 may be identified based on a serial number or other identification number of the input selector 130, for example. [0020] Scattering parameters, also referred to as S-parameters, for each of the ports of the multiplexer 132 may be stored along with the calibration parameters 140 or in the separate memory 136 to assist the test and measurement instrument 160 in de-embedding the input selector 130 from the signal of the DUT 110. The through path and isolation path S-parameters are stored for each of the connectors 141, 143, 145, 147 of the multiplexer 132 as well as the port 162 of the measurement instrument 160. In some examples, a user may upload or otherwise port their own S-parameters, such as if the user wishes to change the S-parameters or add stress to the system).
Hazzard et al does not explicitly teach test plan comprising planned tests.
Ren et al (US 2021/0342516 A1) teaches, a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract). In another aspect, the disclosure provides a method of training a RL agent, employing a game, to determine routing connections for circuits. In one example, this method includes: (1) observing, by the RL agent, a current routing state between terminal positions of a circuit in a routing environment, (2) providing, from the RL agent to the routing environment, a routing action that changes the current routing state between the terminal positions, wherein the RL agent provides the routing action based on a model for routing nets that is independent of a number of the nets of the circuit, (3) evaluating, by the routing environment, the routing action, and (4) providing, from the routing environment to the RL agent, one or more reward based on the evaluating [0006]. In contrast to the limitations of existing methods, the disclosure provides a general solution to determining connections between terminals of circuits. Disclosed herein is a machine learning (ML) based method that uses reinforcement learning (RL), such as deep RL, to determine routing of connections for circuits. The disclosed ML method leverages RL to optimize routing of circuit connections using a game process. In the method, an agent is trained to complete a task of routing connections for a circuit within a routing environment. The agent sends routing actions to the routing environment and in return receives observations and rewards from the routing environment. The rewards provide a measurement of the success of a routing action with respect to completing a routing task within the routing environment [0020]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 19, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 18.
Hazzard et al (US 2022/0390513 A1) further teaches, ([0014] The multiplexer 132 of the input selector 130 provides a selection function to the user. In other words, the user controls the multiplexer 132 of the input selector 130 to choose any of the four inputs to the multiplexer as the desired output of the input selector. The selected signal from the multiplexer 132 is amplified by amplifier 134 before sending it through an RF connection, such as a cable 152, to a test port 162 of the measurement instrument 160. Although in FIG. 1 the amplifier 134 includes only a single output being sent to the measurement instrument 160, in other embodiments the amplifier 134 may send more than one signal to the measurement instrument 160 for testing. For example, the amplifier 134 may be a differential amplifier that sends a differential signal over two separate paths to the measurement instrument 160. In other embodiments the input selector 130 may include multiple amplifiers 134, which each send an amplified signal to the measurement instrument 160. For example the input selector 130 may be controlled to simultaneously send the A input of both Lanes 0 and Lane 1, as separate signals routed from the multiplexer 132 through separate amplifiers 134, to the measurement instrument 160. Or, the input selector 130 may be controlled to simultaneously send both the A and B inputs of Lane 1 out of the multiplexer 132 and individual amplifiers 134, as separate signals, to the measurement instrument 160).
Hazard et al fails to explicitly teach, wherein the system comprises an output interface connected with the processing circuit, and wherein the processing circuit is configured to return a scheduling plan via the output interface, which comprises an optimized order of the planned tests.
Ren et al (US 2021/0342516 A1) teaches, a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract). [0006] In another aspect, the disclosure provides a method of training a RL agent, employing a game, to determine routing connections for circuits. In one example, this method includes: (1) observing, by the RL agent, a current routing state between terminal positions of a circuit in a routing environment, (2) providing, from the RL agent to the routing environment, a routing action that changes the current routing state between the terminal positions, wherein the RL agent provides the routing action based on a model for routing nets that is independent of a number of the nets of the circuit, (3) evaluating, by the routing environment, the routing action, and (4) providing, from the routing environment to the RL agent, one or more reward based on the evaluating. [0020] In contrast to the limitations of existing methods, the disclosure provides a general solution to determining connections between terminals of circuits. Disclosed herein is a machine learning (ML) based method that uses reinforcement learning (RL), such as deep RL, to determine routing of connections for circuits. The disclosed ML method leverages RL to optimize routing of circuit connections using a game process. In the method, an agent is trained to complete a task of routing connections for a circuit within a routing environment. The agent sends routing actions to the routing environment and in return receives observations and rewards from the routing environment. The rewards provide a measurement of the success of a routing action with respect to completing a routing task within the routing environment. [0026] The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections.
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Regarding dependent claim 20, Hazzard et al (US 2022/0390513 A1) and Ren et al (US 2021/0342516 A1) teach the system according to claim 11.
Hazzard et al (US 2022/0390513 A1) further teaches, wherein the system comprises an output interface connected with the processing circuit, and wherein the processing circuit is configured to return a cabling plan via the output interface, which provides the optimized cabling.
Hazzard et al (US 2022/0390513 A1) further teaches ([0037] one or more test and measurement instruments to test one or more signal lanes which may come from one or more DUTs. The input selectors of the disclosure can allow for switching between the various signal lanes to one or more test and measurement instruments. The input selectors have known calibration parameters determined at manufacturing or the factory that can be used by the test and measurement instruments to remove any effects of the selectors, their components, and their connection cables. Operation of the input selector may be programmatically controlled, and a user may program a selector to step through all of the connected signal lanes coming from a DUT without requiring any cables to be disconnected or rearranged. This can save hours of manual labor during a testing session).
Hazzard et al does not explicitly teach wherein the processing circuit returns a cabling plan which provides the optimized cabling.
Ren et al further teaches, wherein the processing circuit returns a cabling plan that provides the optimized cabling ([0026] The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections). [0080] The routed nets from the method 700 can be used in a design flow to produce the circuit. For example, the routed nets can be used in the physical design portion of a design flow that includes planning the layout of the different electrical components, placing the electrical components, and determining the routing of the connections between terminals of the electrical components).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
PNG
media_image1.png
725
469
media_image1.png
Greyscale
Regarding independent claim 21, Hazzard et al (US 2022/0390513 A1)) teaches, A method of optimizing cabling of at least one device under test when performing tests on the at least one device under test (paragraph [0002]), wherein the method comprises the steps of: providing at least one device under test (element 110, figure 1, paragraph [0001]) with a plurality of connectors (elements 121, 123, 125, 127, 141, 143, 145 and 147, figure 1, paragraph [0011]), wherein the connectors are capable of supporting different frequencies and/or different signal directions ([0015] Depending on the DUT 110 and the test and measurement instrument 160, signals may travel both to and from each of the DUT 110 and the test and measurement instrument 160), providing at least one testing device with a plurality of test ports (element 162 of the testing instrument 160 as shown in figure 1, although illustrated in FIG. 2 as only having a single output, the multiplexer 232 may include multiple outputs, especially when there are multiple inputs to the multiplexer from the DUT 110, such as eight, sixteen, thirty-two, or more inputs. Increasing the number of outputs of the multiplexer 232 has the effect of increasing the bandwidth, and therefore the testing speed, of the test and measurement system 200 [0025], [0061]), wherein the test
PNG
media_image2.png
480
719
media_image2.png
Greyscale
ports are capable of supporting different frequencies and/or different signal directions (paragraphs [0004], [0015]), wherein the at least one device under test and the testing device are to be connected with each other via a cabling (RF cable 152, figure 1, paragraph [0014]), providing a switch unit that is interposed between the testing device and the device under test, wherein the switch unit has front end ports as well as back end ports (the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130. (paragraph [0017]), wherein the plurality of connectors of the device under test are at least partly connected with the front end ports of the switch unit, whereas the back end ports of the switch unit are connected with the plurality of test ports of the testing device (figures 1 and 2, paragraphs [0019], [0025]), wherein the cabling comprises connecting cables between the plurality of connectors and the front end ports as well as connecting cables between the plurality of test ports and the back end ports (connectors 121, 123, 125 and 127, 141, 143, 145, 147, and cables 131, 133, 135 and 137 ae shown in figure 1, paragraphs [0019], [0020]; although illustrated in FIG. 2 as only having a single output, the multiplexer 232 may include multiple outputs, especially when there are multiple inputs to the multiplexer from the DUT 110, such as eight, sixteen, thirty-two, or more inputs. Increasing the number of outputs of the multiplexer 232 has the effect of increasing the bandwidth, and therefore the testing speed, of the test and measurement system 200 [0025]).
Hazzard et al (US 2022/0390513 A1) further teaches, ([0017] all of the signals from the DUT 110 may be scripted to be tested without any necessity of the user to physically change any cables between the DUT 110 and the measurement instrument 160. Instead, the switching to connect the desired signal of the DUT 110 to the measurement instrument 160 is performed by controlling the multiplexer 132 of the input selector 130; [0019] In some examples, the input selector 130 may include a memory 136 and/or a controller or processor 138. The input selector 130 may be tested during manufacturing or at the factory to measure the effects of including the input selector 130 in the testing system 100 compared to connecting a particular output of the DUT 110 directly to the test port 162 of the measurement instrument 160. Compensating for any negative effects of including the input selector 130, its components including the multiplexer 132 and amplifier 134, and its related cables 131, 133, 135, 137, and 152 is referred to as de-embedding or calibrating. The term calibration parameters in this disclosure refers to any calibration parameter, including de-embed parameters of the input selector 130, that are used to remove effects or impacts of the presence of the input selector 130 and its related cables 131, 133, 135, 137, and 152 in the testing system 100 from any signal measurements made by the test and measurement instrument 160. The calibration parameters may be stored in a particularized calibration parameters memory 140 or in the general memory 136 and sent to the test and measurement instrument 160 during a testing session. In some examples, the calibration parameters may be sent to an analysis device that is remote from the test and measurement instrument 160. The analysis device may collect the calibration parameters as well as data from the test and measurement instrument 160 and provide any processing needed of the data, as will be understood by one skilled in the art. In other examples, the calibration parameters for a particular input selector 130 may be stored in a memory 163 located in the test and measurement instrument 160 or retrieved from remote storage, such as cloud storage and sent to the input selector 130 to be stored in its calibration parameters storage 140. The particular calibration parameters used to de-embed the effects of the input selector 130 may be identified based on a serial number or other identification number of the input selector 130, for example. [0020] Scattering parameters, also referred to as S-parameters, for each of the ports of the multiplexer 132 may be stored along with the calibration parameters 140 or in the separate memory 136 to assist the test and measurement instrument 160 in de-embedding the input selector 130 from the signal of the DUT 110. The through path and isolation path S-parameters are stored for each of the connectors 141, 143, 145, 147 of the multiplexer 132 as well as the port 162 of the measurement instrument 160. In some examples, a user may upload or otherwise port their own S-parameters, such as if the user wishes to change the S-parameters or add stress to the system).
Hazzard et al does not explicitly teach wherein the processing circuit considers the at least one device under test, the at least one testing device as well as a test plan comprising planned tests when optimizing the cabling between the at least one device under test and the at least one testing device
Ren et al (US 2021/0342516 A1) teaches, a general solution for determining connections between terminals of various types of circuits using machine learning (ML). A ML method that uses reinforcement learning (RL), such as deep RL, to determine and optimize routing of circuit connections using a game process is provided. In one example a method of determining routing connection includes: (1) receiving a circuit design having known terminal groups, (2) establishing terminal positions for the terminal groups in a routing environment, and (3) determining, by the RL agent, routes of nets between the known terminal groups employing a model that is independent of a number of the nets of the circuit. A method of creating a model for routing nets using RL, a method of employing a game for training a RL agent to determine routing connections, and a RL agent for routing connections of a circuit are also disclosed (abstract). In another aspect, the disclosure provides a method of training a RL agent, employing a game, to determine routing connections for circuits. In one example, this method includes: (1) observing, by the RL agent, a current routing state between terminal positions of a circuit in a routing environment, (2) providing, from the RL agent to the routing environment, a routing action that changes the current routing state between the terminal positions, wherein the RL agent provides the routing action based on a model for routing nets that is independent of a number of the nets of the circuit, (3) evaluating, by the routing environment, the routing action, and (4) providing, from the routing environment to the RL agent, one or more reward based on the evaluating [0006]. In contrast to the limitations of existing methods, the disclosure provides a general solution to determining connections between terminals of circuits. Disclosed herein is a machine learning (ML) based method that uses reinforcement learning (RL), such as deep RL, to determine routing of connections for circuits. The disclosed ML method leverages RL to optimize routing of circuit connections using a game process. In the method, an agent is trained to complete a task of routing connections for a circuit within a routing environment. The agent sends routing actions to the routing environment and in return receives observations and rewards from the routing environment. The rewards provide a measurement of the success of a routing action with respect to completing a routing task within the routing environment [0020]). The routing environment 110 is a game environment that represents a grid space for connecting terminal groups of the circuit's nets. The routing environment can be a grid-based routing environment with three layers and a three dimensional grid, such as represented in FIG. 3. The three layers can be a terminal layer, a horizontal layer, and a vertical layer. The routing environment 110 will route nets according to routing actions generated by the RL agent 120. The routing environment 110 evaluates the new routing action and sends the RL agent 120 at least one reward. The routing environment 110 can evaluate the new routing action by determining, for example, if the routing action completes a connection between a terminal group, if the routing action adds a segment for completing the connection, if the routing action complies with DRC, etc. The routing environment 110 produces rewards based on a current routing state, i.e., a game state, and the action, such as, for example, performed via a classic agent-environment loop. The rewards generated can include three different types. One type of reward indicates whether a routing action is illegal or not. These are negative rewards to prevent the agent from creating illegal actions. Another type of reward indicates whether the new routing action created a connected routing segment, i.e. the new routing action is connected to a terminal. These are examples of positive rewards to encourage the RL agent 120 to take routing actions that connect with existing terminals. A third type of reward is net routing rewards, which are given when a net is fully routed. The amount of that reward can be associated with the routed wirelength. A shorter wire length can result in larger rewards to encourage the RL agent to optimize lengths of wires, i.e., connections (paragraph [0026]).
Therefore it would have been obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention, to have modified the teachings of Hazzard by providing for optimizing the routing of connections between circuit elements as taught by Ren et al (paragraph [0026]).
One of the ordinary skill in the art would have been motivated to make such a modification to adopt the teachings of optimizing the length of wires i.e. connections, so that the connection between the device under test and the testing device can be optimized, as taught by Ren et al (paragraph [0048], [0049]).
Closest Prior art
7. The following relevant prior art of record is not cited in the office action.
Peters et al (US 2024/0019484 A1) teaches, An apparatus includes an interface to a device-under-test (DUT), an interface to a rack of cable holders, and a control circuit. Each cable holder is configured to raise or lower respective cable. The control circuit is configured to determine an internal configuration of the DUT and, based upon an internal configuration of the DUT, identify a plurality of cables to be used in testing the DUT. The control circuit is configured to cause the rack of cable holders to actuate a plurality of the cable holders associated with the identified plurality of cables to raise or lower the cable holders associated with the identified plurality of cables.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SURESH RAJAPUTRA whose telephone number is (571) 270-0477. The examiner can normally be reached between 8:00 AM - 5:00 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, EMAN ALKAFAWI can be reached on 571-272-4448. 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.
/SURESH K RAJAPUTRA/Examiner, Art Unit 2858
/EMAN A ALKAFAWI/Supervisory Patent Examiner, Art Unit 2858 2/6/2026