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, 25 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.
Argument: Applicant argued “However, the Office Action has not shown Ren to teach or suggest "receiving a first message indicating one or more life cycle management trigger conditions for a channel between the UE and a network entity," as recited in independent claim 22.
Response: As stated in claim 21 rejection, WO089 discloses ([0004], [0114]) “if a ML model is performing relatively poorly … “, wherein from [0004] and [0114], it is obvious that the performance is in relation to the UE – Base station channel performance. This is further disclosed by WO089 in [0056], where it is stated:
“However, some operating conditions at the UE may negatively affect the performance of an ML model. For example, based on the actual inputs to the ML model during UE operation ( e.g., channel measurements, signal measurements, or other current operating conditions), the ML model may degrade the performance of the UE (e.g., as compared to an alternative mode in which the UE does not use the ML model).”
[0057] discloses the machine learning model may then be updated.
Hence, as recited in claim 21, the lifecycle is managed based on the trigger and report.
Examiner is using Yang et al. (US 20250081033 A1) to address the new limitation in claims 1, 25.
Claim Rejections - 35 USC § 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.
Claims 1 - 13, 25 - 28 are rejected under 35 U.S.C. 103 as being unpatentable over Zhu et al. (US 20230100253 A1; which has been provided in the International Search Report) in view of Yang et al. (US 20250081033 A1).
Regarding claim 1, Zhu discloses a network entity (Fig. 1, base stations 110; [0040]), comprising:
one or more memories storing processor-executable code (Fig. 3, element 318; [0061] – [0062]);
and one or more processors coupled with the one or more memories and individually or collectively operable to execute the code (Fig. 3, elements 302, 304, 306, 308; [0061] – [0062]) to cause the network entity to:
obtain a first message indicating one or more machine learning capabilities of a user equipment (UE) (Fig. 8, step 1; [0095] discloses “At time 1, the UE 120 transmits, to the base station 110, UE capability information indicating at least one radio capability of the UE and at least one machine learning capability of the UE”; [0095] – [0097]);
obtain, from a machine learning service, a second message indicating one or more machine learning service capabilities of the machine learning service ([0100] – [0101] discloses selecting atleast one machine learning model; wherein the capability would be all the models it is capable of selecting from);
and output, to the UE, a control message indicate one or more cross- node machine learning configurations based at least in part on the one or more machine learning capabilities of the UE and the one or more machine learning service capabilities of the machine learning service, wherein the one or more cross-node machine learning configurations configure one or more machine learning functions of the UE for use with the machine learning service ([0102] discloses “Thereafter, the base station 110 (e.g., via the CU-CP 712) transmits, to the UE 120, machine learning configuration information based on the UE capability information”; [0102]. As per Applicant’s specification [0046], last sentence: “In any case, the deployment of AI/ML functions across multiple devices and/or entities may be referred to as cross-node inference, where AI/ML-related inference is performed by the UE as well as the AI/ML service and/or RAN”. Hence, since Zhu discloses machine learning deployed across a UE and network entity, this is interpreted as cross-node).
Zhu does not disclose the machine learning service is separate from the network entity.
In the same field of endeavor, however, Yang discloses the machine learning service is separate from the network entity (Fig. 1, MEC device 145 is separate from base station 120; [0023] discloses “MEC network 140 may include one or more MEC devices 145. MEC devices 145 may provide MEC services to UE devices 110. A MEC service may include …. artificial intelligence (AI) accelerator service, machine learning service,…”).
Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to use the method, as taught by Yang, in the system of Zhu because having the machine learning service separate from the network entity ensures that the workload is divided between the station and MEC, hence the base station won’t be overloaded.
Regarding claim 2, Zhu discloses the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
output, to the machine learning service, a service capability request message for the one or more machine learning service capabilities ([0099] - [0100]), wherein the second message indicating the one or more machine learning service capabilities is obtained in response to the service capability request message ([0100] discloses “The CU-XP 716 may then select at least one machine learning model for use in the at least one NNF to perform at least the portion of the machine learning-based wireless communications management procedure. In some cases, the CU-XP 716 may select the at least one machine learning model based, at least in part, on the at least one machine learning capability of the UE…”; wherein “then select” implies after the request message is received), and wherein the one or more machine learning service capabilities comprise capabilities that are compatible with the UE based at least in part on the one or more machine learning capabilities of the UE ([0100] discloses “In some cases, the CU-XP 716 may select the at least one machine learning model based, at least in part, on the at least one machine learning capability of the UE”).
Regarding claim 3, Zhu discloses the service capability request message comprises a set of UE identifiers including an identifier of the UE, an indication of the one or more machine learning capabilities of to the UE, or any combination thereof ([0100] discloses UE context information, UE type).
Regarding claim 4, Zhu discloses wherein, to obtain the second message indicating the one or more machine learning service capabilities, the one or more processors are individually or collectively operable to execute the code to cause the network entity (as in claim 1) to:
obtain, from the machine learning service, an announcement of the one or more machine learning service capabilities, the announcement comprising the second message ([0101] discloses “a context setup response message”; wherein the announcement is the context setup response message).
Regarding claim 5, Zhu discloses wherein the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
select, based at least in part on the one or more machine learning capabilities of the UE and the one or more machine learning service capabilities of the machine learning service, the one or more machine learning functions ([0100] discloses “In some cases, the CU-XP 716 may select the at least one machine learning model based, at least in part, on the at least one machine learning capability of the UE”);
output, to the machine learning service, a configuration request message requesting one or more cross-node machine learning configurations based at least in part on the selecting ([0100]);
and obtain, from the machine learning service, a configuration response message that indicates the one or more cross-node machine learning configurations, the one or more machine learning functions, a set of UE identifiers including an identifier of the UE, or any combination thereof, wherein the control message indicating one or more cross-node machine learning configurations is based at least in part on obtaining the configuration response message ([0100]; wherein ID is UE context information, UE type).
Regarding claim 6, Zhu discloses wherein the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
obtain a third message indicating completion of the one or more cross- node machine learning configurations by the UE in response to the control message ([0108]).
Regarding claim 7, Zhu discloses wherein the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
output, to the machine learning service, a fourth message indicating that the one or more cross-node machine learning configurations have been configured by the UE ([0108]).
Regarding claim 8, Zhu discloses wherein the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
output a second control message comprising an indication to activate the one or more machine learning functions based at least in part on the one or more cross- node machine learning configurations ([0109] – [0110] discloses activation signal).
Regarding claim 9, Zhu discloses wherein the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
output, to the machine learning service, an activation request message indicating a set of UE identifiers including an identifier of the UE, the one or more machine learning functions ([0109] discloses activation signal for machine learning), or any combination thereof.
Zhu does not explicitly disclose:
and obtain, from the machine learning service, an activation acknowledgment message from the machine learning service in response to the activation request message, the activation acknowledgment message indicating the set of UE identifiers including the identifier of the UE, the one or more machine learning functions, or any combination thereof, wherein the second control message is output based at least in part on the activation acknowledgment message.
However, sending an acknowledgement message is obvious to try (Rationales for Obviousness (MPEP 2143, Rationale E)) and can be easily done similar to acknowledgement messages shown in Fig. 9 (e.g. step 6).
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to send an acknowledgement message, as this would ensure that the activation message has been received.
Claim 10 is similarly analyzed as claim 9, with the same operations performed in the reverse direction. Hence, one of ordinary skill in the art can do the same as in claim 9 for the same reasons.
Claim 11 is similarly analyzed as claims 9 and 10. Claim 11 merely recites obtaining activation request message, outputting a response and obtaining an acknowledgement. These are similar operations to those in claim 9 and can be performed by any of the network entity, UE, or machine learning service.
Regarding claim 12, Zhu discloses:
a centralized unit control plane (CU-CP) and/or centralized unit machine learning plane (CU-XP) may decide to configure a network model for inference and/or training ([0113]).
The following claim limitations are an obvious variation of the above (Rationales for Obviousness (MPEP 2143, Rationale & F)):
obtain a sixth message comprising UE inference input data associated with the one or more machine learning functions;
output, to the machine learning service, a service data request message comprising the UE inference input data;
and obtain, from the machine learning service, a service data response message indicating machine learning service inference output data associated with the one or more machine learning functions.
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to do the above, as these are essential and well-known steps associated with machine learning/AI that one uses to obtain results from ML/AI.
Regarding claim 13, Zhu discloses:
a centralized unit control plane (CU-CP) and/or centralized unit machine learning plane (CU-XP) may decide to configure a network model for inference and/or training ([0113]).
The following claim limitations are an obvious variation of the above (Rationales for Obviousness (MPEP 2143, Rationale & F)):
output, to the UE, a seventh message comprise the machine learning service inference output data from the machine learning service.
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to do the above, as these are essential and well-known steps associated with machine learning/AI that one uses to obtain results from ML/AI.
Claim 25 is similarly analyzed as claim 1, with claim 25 reciting equivalent method limitations.
Claim 26 is similarly analyzed as claim 2.
Claim 27 is similarly analyzed as claim 3.
Claim 28 is similarly analyzed as claim 4.
Claims 14 - 24, 29 - 30 are rejected under 35 U.S.C. 103 as being unpatentable over Zhu et al. (US 20230100253 A1) in view of Yang et al. (US 20250081033 A1) and further in view of WO2022222089A1 (hereafter, WO089), which have been provided in the International Search Report).
Regarding claim 14, Zhu does not disclose wherein the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
monitor for one or more trigger conditions based at least in part on the UE inference input data, machine learning service inference input data from the machine learning service, or any combination thereof;
and output, to the UE, a third control message comprise an indication to switch or deactivate at least one of the one or more machine learning functions based at least in part on an occurrence of the one or more trigger conditions.
In the same field of endeavor, however, WO089 discloses:
monitor for one or more trigger conditions based at least in part on the UE inference input data, machine learning service inference input data from the machine learning service, or any combination thereof ([0100]);
and output, to the UE, a third control message comprise an indication to switch or deactivate at least one of the one or more machine learning functions based at least in part on an occurrence of the one or more trigger conditions ([0100] discloses “…configure and trigger ML model updating at 350,…”; wherein switch would correspond to updating the model.
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions.
Regarding claim 15, Zhu does not disclose the one or more trigger conditions comprises a measurement of one or more key performance indicators satisfying a key performance indicator threshold.
In the same field of endeavor, however, WO089 discloses the one or more trigger conditions comprises a measurement of one or more key performance indicators satisfying a key performance indicator threshold ([0004] discloses “…if an ML model is performing relatively poorly (e.g., below a performance threshold), … the UE may trigger a transmission of a status report to a base station…”; [0117]).
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions so that better ML models could be used.
Regarding claim 16, Zhu does not disclose the one or more processors are individually or collectively further operable to execute the code to cause the network entity to: output, to the machine learning service, a monitoring report that indicates one or more key performance indicators associated with the one or more machine learning functions.
In the same field of endeavor, however, WO089 discloses the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
output, to the machine learning service, a monitoring report that indicates one or more key performance indicators associated with the one or more machine learning functions ([0116] discloses “For example, the network (e.g., the base station 105-b) may monitor one or more metrics associated with the UE 115-c, the ML model at the UE 115-c, or both. Based on the one or more metrics, the network may trigger an ML model status report. For example, if a performance loss of the UE 115-c satisfies a threshold performance loss, the network may trigger model status reporting from the UE 115-c (e.g., if the performance loss may potentially indicate an ML model failure at the UE 115-c).”
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions so that better ML models could be used.
Regarding claim 17, Zhu does not disclose the one or more trigger conditions are based at least in part on a monitoring report comprising measurements performed by the UE, by the network entity, or any combination thereof.
In the same field of endeavor, however, WO089 discloses the one or more trigger conditions are based at least in part on a monitoring report comprising measurements performed by the UE, by the network entity, or any combination thereof ([0116]; wherein the measurements are the metrics).
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions so that better ML models could be used.
Regarding claim 18, Zhu does not disclose the one or more processors are individually or collectively further operable to execute the code to cause the network entity to: obtain a monitoring report from the UE, the monitoring report comprising one or more key performance indicators associated with the one or more machine learning functions; and output, to the UE, an indication to switch or deactivate the one or more machine learning functions of the UE based at least in part on an occurrence of one or more trigger conditions.
In the same field of endeavor, however, WO089 discloses the one or more processors are individually or collectively further operable to execute the code to cause the network entity to:
obtain a monitoring report from the UE, the monitoring report comprising one or more key performance indicators associated with the one or more machine learning functions ([0116]);
and output, to the UE, an indication to switch or deactivate the one or more machine learning functions of the UE based at least in part on an occurrence of one or more trigger conditions ([0100], as in claim 14 above).
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions so that better ML models could be used.
Regarding claim 19, Zhu does not disclose the one or more processors are individually or collectively further operable to execute the code to cause the network entity to: output, to the machine learning service, a monitoring report that indicates one or more key performance indicators associated with the one or more machine learning functions.
In the same field of endeavor, however, WO089 discloses the one or more processors are individually or collectively further operable to execute the code to cause the network entity to: output, to the machine learning service, a monitoring report that indicates one or more key performance indicators associated with the one or more machine learning functions ([0116] discloses “For example, the network (e.g., the base station 105-b) may monitor one or more metrics associated with the UE 115-c, the ML model at the UE 115-c, or both. Based on the one or more metrics, the network may trigger an ML model status report. For example, if a performance loss of the UE 115-c satisfies a threshold performance loss, the network may trigger model status reporting from the UE 115-c (e.g., if the performance loss may potentially indicate an ML model failure at the UE 115-c).”
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions so that better ML models could be used.
Regarding claim 20, Zhu does not disclose the one or more processors are individually or collectively further operable to execute the code to cause the network entity to: obtain, from the UE, life cycle management control request message comprising a request for life cycle management control signaling, the life cycle management control signaling comprising an indication for activating, deactivating, switching, a default configuration, or any combination thereof, of the one or more cross- node machine learning configurations; output, to the machine learning service, a first life cycle management control message comprise an indication of request for the life cycle management control signaling in response to the life cycle management control request message; and output, to the UE, a second life cycle management control message comprising and indication of the life cycle management control signaling indicated by the machine learning service.
In the same field of endeavor, however, WO089 discloses activating, deactivating, switching, a default configuration, or any combination thereof ([0100] discloses updating) and outputting ([0116], as in claim 16 above). Outputting the results to various network elements is an obvious variation of the above (Rationales for Obviousness (MPEP 2143, Rationale F)).
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for other network elements to be aware of a ML model degradation and then updating ML models based so that better ML models could be used.
Regarding claim 21, Zhu does not disclose obtaining the life cycle management control request message is based at least in part on a monitoring report from the UE, the network entity, or a combination thereof.
In the same field of endeavor, however, WO089 discloses obtaining the life cycle management control request message is based at least in part on a monitoring report from the UE, the network entity, or a combination thereof ([0004] discloses “…if an ML model is performing relatively poorly (e.g., below a performance threshold), … the UE may trigger a transmission of a status report to a base station…”; [0117]; [0056] discloses poor channel conditions can cause a trigger; [0057] discloses machine learning model is then updated). Hence, the lifecycle is managed based on the trigger and report.
Therefore, it would have been obvious to one having ordinary skill in the art, at the time the invention was filed, to use the method, as taught by WO089in the system of Zhu because this would allow for updating ML models based on trigger functions so that better ML models could be used.
Claim 22 is similarly analyzed as in claims 14, 17, 20, 21 wherein the 1st limitation of claim 22 is similarly analyzed as in claims 14, 17 (i.e. trigger detected and report sent with trigger information).
Claim 23 is similarly analyzed as claims 14, 17, 21.
Claim 24 is similarly analyzed as claim 15.
Claim 29 is similarly analyzed as claim 22, with claim 29 reciting equivalent method limitations.
Claim 30 is similarly analyzed as claim 23.
Other Prior Art Cited
The prior art made of record and not relied upon is considered pertinent to the applicant’s disclosure.
The following patents/publications are cited to further show the state of the art with respect to using ML in network systems:
Bai et al. (US 20220248312 A1) discloses Method for Establishing Communication Between User Equipment and Network Device for Selecting Machine Learning Modules in Communication System, Involves Selecting Configuration of Machine Learning Module Based on Selection Parameter.
Conclusion
THIS ACTION IS MADE FINAL. 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.
Contact Information
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ADOLF DSOUZA whose telephone number is (571)272-1043. The examiner can normally be reached Mon - Fri 9 AM - 5 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, Chieh M Fan can be reached at 571-272-3042. 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.
/ADOLF DSOUZA/Primary Examiner, Art Unit 2632