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 .
Response to Arguments
Applicant’s arguments with respect to claims 1-15 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. The scope of the claims have changed due to the amendments made in independent claim 1 for example.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-5, 9, and 12-15 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Hajri et al. (US 2022/0386292).
Regarding claim 1, Hajri discloses a method performed by a user equipment (UE), the method comprising: acquiring information regarding an artificial intelligence/machine learning (AI/ML) model ([0069], “At 302, the UE receives a radio resource control configuration including at least one CSI reporting configuration for which prediction is configured/enabled and multiple prediction windows for one or multiple CSI quantities. At 304, the UE receives downlink control information triggering CSI reporting containing, at least, one CSI quantity for which prediction is enabled/configured.”); monitoring the AI/ML model ([0069], “At 306, the UE receives downlink reference signals for a channel and/or interference measurements. At 308, the UE computes CSI quantity values and/or CSI quantity prediction models, based on downlink reference measurements, the CSI reporting configuration and configured prediction windows.”); and performing a channel state information (CSI) report procedure in an AI/ML-based CSI report mode or a non-AI/ML-based CSI report mode ([0069], “At 310, for CSI quantities for which CSI prediction is enabled/configured, the UE selects one or multiple prediction windows, based on its channel state and/or position and/or velocity and/or a network/UE/specification-defined prediction performance metric. As further described at 310, selected prediction windows may be quantity-specific or common for all or a subset of CSI quantities. At 312, the UE transmits triggered CSI report(s) in uplink control information.”), based on a validity of the monitored AI/ML model wherein ([0069], “As further indicated at 312, the UE feeds back one or multiple prediction window indicators and, for CSI quantities for which prediction is enabled/activated, either a model valid for the selected prediction window or multiple quantized values of the CSI quantity spanning the selected prediction window.”), based on the monitored AI/ML model being invalid ([0061], “Early learning termination. In case the gNB detects a considerable change in the UE channel conditions, e.g. based on SRS, DMRS or TRS, the gNB may instruct the UE, via dynamic downlink signaling, to terminate its learning process.”), the UE switches from the AI/ML-based CSI report mode to the non-AI/ML-based CSI report mode as a fallback operation for the invalid AI/ML model ([0061], “Depending on a specified rule, the UE may reset its model to learn from scratch based on new samples, fallback to conventional CSI operation without prediction until receiving a prediction reactivation command from the network, or adapt its learning rate autonomously or based on an indicated rate or rate offset from the network.”).
Regarding claim 2, Hajri discloses the method of claim 1, wherein the UE determines that the AI/ML model is invalid based on a performance value of the monitored AI/ML model being less than a threshold ([0061], “Early learning termination. In case the gNB detects a considerable change in the UE channel conditions, e.g. based on SRS, DMRS or TRS, the gNB may instruct the UE, via dynamic downlink signaling, to terminate its learning process.”).
Regarding claim 3, Hajri discloses the method of claim 1, wherein performing the CSI report procedure comprises: transmitting a report including at least one of a result of the monitoring of the AI/ML model or a channel quality report based on the AI/ML model, to a base station (BS) ([0069], “At 310, for CSI quantities for which CSI prediction is enabled/configured, the UE selects one or multiple prediction windows, based on its channel state and/or position and/or velocity and/or a network/UE/specification-defined prediction performance metric. As further described at 310, selected prediction windows may be quantity-specific or common for all or a subset of CSI quantities. At 312, the UE transmits triggered CSI report(s) in uplink control information.”).
Regarding claim 4, Hajri discloses the method of claim 3, wherein the validity of the AI/ML model is determined by the BS based on the report ([0061], “In case the gNB detects a considerable change in the UE channel conditions, e.g. based on SRS, DMRS or TRS, the gNB may instruct the UE, via dynamic downlink signaling, to terminate its learning process. Depending on a specified rule, the UE may reset its model to learn from scratch based on new samples, fallback to conventional CSI operation without prediction until receiving a prediction reactivation command from the network, or adapt its learning rate autonomously or based on an indicated rate or rate offset from the network.”).
Regarding claim 5, Hajri discloses the method of claim 1, further comprising: receiving information regarding a threshold for determining the validity of the AI/ML model from a base station (BS) ([0057], “At 204, the UE receives downlink control information indicating a prediction window and triggering CSI reporting containing at least one CSI quantity for which prediction is enabled/configured. At 206, the UE receives downlink reference signals for channels and/or interference measurements. At 208, the UE computes CSI quantity values and/or CSI quantity prediction models, based on downlink reference measurements, the CSI reporting configuration and the indicated prediction window.”).
Regarding claim 9, Hajri discloses the method of claim 1, wherein the UE determines that the AI/ML model is invalid based on an operating time of the AI/ML model exceeding a maximum time duration ([0040], “If a prediction window indicates a single time unit offset, depending on a specified or configured rule, the time unit offset may be indicating one of the following (1-3): 1) An upper bound of the prediction time interval, which refers to the highest duration of the time unit offset for which the predicted CSI quantity should be valid, or the maximum duration of the offset for the output of a CSI prediction model.”).
Regarding claim 12, Hajri discloses a non-transitory computer-readable medium having recorded thereon a program for executing the method of claim 1 ([0084], “FIG. 7 is an example apparatus 700, which may be implemented in hardware, configured to implement the examples described herein. The apparatus 700 comprises at least one processor 702 (an FPGA and/or CPU), at least one non-transitory or transitory memory 704 including computer program code 705”).
Regarding claim 13, Hajri discloses a device comprising: a memory configured to store instructions; and a processor configured to perform operations by executing the instructions, wherein the operations of the processor comprise ([0084], “The apparatus 700 comprises at least one processor 702 (an FPGA and/or CPU), at least one non-transitory or transitory memory 704 including computer program code 705, wherein the at least one memory 704 and the computer program code 705 are configured to, with the at least one processor 702, cause the apparatus 700 to implement circuitry, a process, component, module, or function (collectively control 706) to implement CSI prediction configuration and control.”): acquiring information regarding an artificial intelligence/machine learning (AI/ML) model ([0069], “At 302, the UE receives a radio resource control configuration including at least one CSI reporting configuration for which prediction is configured/enabled and multiple prediction windows for one or multiple CSI quantities. At 304, the UE receives downlink control information triggering CSI reporting containing, at least, one CSI quantity for which prediction is enabled/configured.”); monitoring the AI/ML model ([0069], “At 306, the UE receives downlink reference signals for a channel and/or interference measurements. At 308, the UE computes CSI quantity values and/or CSI quantity prediction models, based on downlink reference measurements, the CSI reporting configuration and configured prediction windows.”); and performing a channel state information (CSI) report procedure in an AI/ML-based CSI report mode or a non-AI/ML-based report mode ([0069], “At 310, for CSI quantities for which CSI prediction is enabled/configured, the UE selects one or multiple prediction windows, based on its channel state and/or position and/or velocity and/or a network/UE/specification-defined prediction performance metric. As further described at 310, selected prediction windows may be quantity-specific or common for all or a subset of CSI quantities. At 312, the UE transmits triggered CSI report(s) in uplink control information.”), based on a validity of the monitored AI/ML model wherein ([0069], “As further indicated at 312, the UE feeds back one or multiple prediction window indicators and, for CSI quantities for which prediction is enabled/activated, either a model valid for the selected prediction window or multiple quantized values of the CSI quantity spanning the selected prediction window.”), based on the monitored AI/ML model being invalid ([0061], “Early learning termination. In case the gNB detects a considerable change in the UE channel conditions, e.g. based on SRS, DMRS or TRS, the gNB may instruct the UE, via dynamic downlink signaling, to terminate its learning process.”) device switches from the AI/ML-based CSI report mode to the non-AI/ML-based CSI report mode as a fallback operation for the invalid AI/ML model ([0061], “Depending on a specified rule, the UE may reset its model to learn from scratch based on new samples, fallback to conventional CSI operation without prediction until receiving a prediction reactivation command from the network, or adapt its learning rate autonomously or based on an indicated rate or rate offset from the network.”).
Regarding claim 14, Hajri discloses a method performed by a base station (BS), the method comprising: transmitting, to a user equipment (UE) ([0016], “The UE 110 communicates with RAN node 170 via a wireless link 111.”), information regarding an artificial intelligence/machine learning (AI/ML) model ([0069], “At 302, the UE receives a radio resource control configuration including at least one CSI reporting configuration for which prediction is configured/enabled and multiple prediction windows for one or multiple CSI quantities. At 304, the UE receives downlink control information triggering CSI reporting containing, at least, one CSI quantity for which prediction is enabled/configured.”); and receiving, from the UE, a channel state information (CSI) report in an AI/ML-based CSI reception mode or a non-AI/ML-based CSI reception mode ([0069], “At 310, for CSI quantities for which CSI prediction is enabled/configured, the UE selects one or multiple prediction windows, based on its channel state and/or position and/or velocity and/or a network/UE/specification-defined prediction performance metric. As further described at 310, selected prediction windows may be quantity-specific or common for all or a subset of CSI quantities. At 312, the UE transmits triggered CSI report(s) in uplink control information.”), based on a validity of the AI/ML model wherein ([0069], “As further indicated at 312, the UE feeds back one or multiple prediction window indicators and, for CSI quantities for which prediction is enabled/activated, either a model valid for the selected prediction window or multiple quantized values of the CSI quantity spanning the selected prediction window.”), based on the AI/ML model being invalid ([0061], “Early learning termination. In case the gNB detects a considerable change in the UE channel conditions, e.g. based on SRS, DMRS or TRS, the gNB may instruct the UE, via dynamic downlink signaling, to terminate its learning process.”) the BS switches ([0069], “FIG. 3 shows a UE procedure 300 with UE-centric prediction window selection. Method 300 may be performed by the UE 110 as shown in FIG. 1, and in part by the RAN node 170 and/or network element(s) 190 shown in FIG. 1.”) from the AI/ML-based CSI reception mode to the non-AI/ML-based CSI reception mode as a fallback operation for the invalid AI/ML model ([0061], “Depending on a specified rule, the UE may reset its model to learn from scratch based on new samples, fallback to conventional CSI operation without prediction until receiving a prediction reactivation command from the network, or adapt its learning rate autonomously or based on an indicated rate or rate offset from the network.”).
Regarding claim 15, Hajri discloses a base station (BS) comprising: a transceiver; and a processor configured to control the transceiver ([0018], “The RAN node 170 includes one or more processors 152, one or more memories 155, one or more network interfaces (N/W I/F(s)) 161, and one or more transceivers 160 interconnected through one or more buses 157. Each of the one or more transceivers 160 includes a receiver, Rx, 162 and a transmitter, Tx, 163.”) to transmit, to a user equipment (UE) ([0016], “The UE 110 communicates with RAN node 170 via a wireless link 111.”), information regarding an artificial intelligence/machine learning (AI/ML) model ([0069], “At 302, the UE receives a radio resource control configuration including at least one CSI reporting configuration for which prediction is configured/enabled and multiple prediction windows for one or multiple CSI quantities. At 304, the UE receives downlink control information triggering CSI reporting containing, at least, one CSI quantity for which prediction is enabled/configured.”); and receive, from the UE, a channel state information (CSI) report in an AI/ML-based CSI reception mode or a non-AI/ML-based CSI reception mode ([0069], “At 310, for CSI quantities for which CSI prediction is enabled/configured, the UE selects one or multiple prediction windows, based on its channel state and/or position and/or velocity and/or a network/UE/specification-defined prediction performance metric. As further described at 310, selected prediction windows may be quantity-specific or common for all or a subset of CSI quantities. At 312, the UE transmits triggered CSI report(s) in uplink control information.”), based on the validity of the AI/ML model wherein ([0069], “As further indicated at 312, the UE feeds back one or multiple prediction window indicators and, for CSI quantities for which prediction is enabled/activated, either a model valid for the selected prediction window or multiple quantized values of the CSI quantity spanning the selected prediction window.”), based on the AI/ML model being invalid ([0061], “Early learning termination. In case the gNB detects a considerable change in the UE channel conditions, e.g. based on SRS, DMRS or TRS, the gNB may instruct the UE, via dynamic downlink signaling, to terminate its learning process.”) the BS switches ([0069], “FIG. 3 shows a UE procedure 300 with UE-centric prediction window selection. Method 300 may be performed by the UE 110 as shown in FIG. 1, and in part by the RAN node 170 and/or network element(s) 190 shown in FIG. 1.”) from the AI/ML-based CSI reception mode to the non-AI/ML-based CSI reception mode as a fallback operation for the invalid AI/ML model ([0061], “Depending on a specified rule, the UE may reset its model to learn from scratch based on new samples, fallback to conventional CSI operation without prediction until receiving a prediction reactivation command from the network, or adapt its learning rate autonomously or based on an indicated rate or rate offset from the network.”).
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 6 and 8 are rejected under 35 U.S.C. 103 as being unpatentable over Hajri et al. (US 2022/0386292) in view of Jia et al. (US 2024/0323741).
Regarding claim 6, Hajri does not disclose the model being invalid due to model output variance. Jia discloses the method of claim 1, wherein the UE ([0076], "the first communication device may be a terminal") determines that the AI/ML model is invalid ([0422], "determining, by the first communication device, validity indication information of the target artificial intelligence model based on the comparison result.") based on a variance of output ([0547], "the performance indication information obtained through statistics for the plurality of continuous measurement time points or the plurality of continuous measurement positions" and [0549], "communication performance statistical indicator, for example a statistical value of a communication performance indicator such as SNR, SINR, packet error rate, interruption probability, or cell/beam switching probability, for example, mean, variance") of the AI/ML model exceeding a threshold ([0420], "determining, by the first communication device, a comparison result between performance of the target artificial intelligence model and a performance threshold"). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Hajri in view of Jia to have the model being invalid due to model output variance. The motivation would have been to reduce errors of the output (e.g., Jia [0067]).
Regarding claim 8, Hajri does not disclose the monitoring periodicity for the model. Jia discloses the method of claim 1, wherein the monitoring of the AI/ML model is based on a preconfigured monitoring periodicity ([0504], "In some embodiments, the first communication device may periodically perform the target measurement based on the first measurement period, for example, a measurement window runs once every 10 ms."). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Hajri in view of Jia to have the monitoring periodicity for the model. The motivation would have been to reduce errors of the output (e.g., Jia [0067]).
Claims 7 and 10-11 are rejected under 35 U.S.C. 103 as being unpatentable over Hajri et al. (US 2022/0386292) in view of Lo et al. (US 2023/0337036; supported by US Provisional 63332579).
Regarding claim 7, Hajri does not disclose the model being invalid being on channel quality being lower than the non-AL/ML model based report mode. Lo discloses the method of claim 1, wherein the UE determines that the AI/ML model is invalid ([0192], “In one embodiment, a rule is pre-determined to prevent error propagation that can happen in a process of the CSI report for differential CSI prediction, e.g., when a UE is configured to report its prediction of the difference between the current CSI observation and the next CSI prediction. The rule follows at least one of the following examples.”) based on channel quality in the AI/ML-based CSI report mode being lower by a threshold than channel quality in the non-AI/ML-based CSI report mode ([0193], “In one example, a parameter to enable the UE to count the occasion (a) of the CSI report for differential CSI prediction can be configured to the UE. Once it is enabled, the UE counts the occasion of the CSI report for differential CSI prediction and reports it to the NW as a part of the CSI report. If the reported value of a is different from the value at the NW, the NW may reset the process for differential CSI prediction to the legacy CSI process (i.e., fallback mode to perform the legacy CSI report).”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Hajri in view of Lo to have the model being invalid being on channel quality being lower than the non-AL/ML model based report mode. The motivation would have been to improve performance (e.g., Lo [0084]).
Regarding claim 10, Hajri does not disclose the updating of the model in the non-AI/ML-based mode. Lo discloses the method of claim 1, wherein the UE updates the invalid AI/ML model in the non-AI/ML-based CSI report mode ([0198], “In one example, a NW can configure a UE to report a measure of uncertainty in its differential CSI prediction. Reporting of this uncertainty metric can be configured via DCI, MAC-CE, or RRC. If the reported measure of uncertainty exceeds a pre-defined threshold, in one example, a NW can configure a UE to convey legacy CSI reports. In another example, a NW can configure a UE to apply another CSI prediction method (e.g., a conventional CSI predictor, an AI-based full CSI predictor that predicts the next CSI observation, another AI-based differential CSI predictor, etc.).”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Hajri in view of Lo to have the updating of the model in the non-AI/ML-based mode. The motivation would have been to improve performance (e.g., Lo [0084]).
Regarding claim 11, Hajri does not disclose the updating of invalid model from BS. Lo discloses the method of claim 10, wherein the updating of the invalid AI/ML model comprises acquiring information regarding an updated AI/ML model from a base station (BS) ([0197], “If the reported differential CSI prediction error exceeds a pre-defined threshold, in one example, a NW can configure a UE to convey legacy CSI reports. In another example, a NW can configure a UE to apply another CSI prediction method (e.g., a conventional CSI predictor, an AI-based full CSI predictor that predicts the next CSI observation, another AI-based differential CSI predictor, etc.).”). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Hajri in view of Lo to have the updating of invalid model from BS. The motivation would have been to improve performance (e.g., Lo [0084]).
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 nonprovisional extension fee (37 CFR 1.17(a)) 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Nick A Sundara whose telephone number is (571)272-6749. The examiner can normally be reached M-TH 7:30-5:30 EST.
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, Jae Y. Lee can be reached at (571) 270-3936. 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.
/NICK ANON SUNDARA/Examiner, Art Unit 2479 /JAE Y LEE/Supervisory Patent Examiner, Art Unit 2479