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 .
This action is in response to applicant’s remarks/arguments filed on 07/13/2026. Currently, claims 1-30 are pending. This action is made FINAL.
Response to Arguments
Applicant's arguments/remarks filed 07/13/2026 have been fully considered but they are not persuasive.
Regarding claims 1 and 12, Applicant’s argues that “[n]either Li nor Ly, alone or in combination, discloses an indication that activates transmission of ML data” and states that “Ly discloses an indication within a PDU session establishment request to enable AI/ML communications for that session – this establish a communication channel, not activate transmission of specific ML data” and “The difference between Ly’s teaching and the claimed feature is the different between (a) establishing a session capable of carrying AI/ML traffic and (b) activating transmission of specific ML data associated with an MLFN”. The Examiner respectfully disagrees. While neither Li or Ly, alone, discloses the claimed limitation in question, the combination of Li and Ly discloses it.
Applicant treats Ly’s indication as establishing only a channel; however, Ly links the request containing the indication to the acceptable of the session for transmitting or receiving AI or ML traffic and subsequent communication of that traffic (par [0096] and [0228] of Ly); thus, the request initiates establishment of the communication path used for the AI or ML traffic including model training results and training data. Li discloses the update request and the responsive MachineLearningModeUpdate with the corresponding service type and parameter updates use for model update and iteration (service associated ML data) (par [0020], [0044] and [0060] of Li). Incorporating Ly’s teaching into Li’s teaching, Li’s update request would initiate establishment of the communication path for Li’s requested service associated parameter update (i.e. service associated ML data) over an accepted session designated for AI or ML traffic communication (see par [0228] of Ly). Therefore, the combination of Li and Ly discloses the claimed limitation in question above and the arguments are not persuasive. The rejection is maintained, see detailed rejection for claims 1 and 12 below with more clarifications.
Regarding claims 24 and 27, Applicant’s argues that the combination of Li and Garcia Rodriguez’s does not teach the limitation “receive an indication from the network entity to activate transmission of the ML data associated with the at least one MLFN” by stating that Garcia Rodriguez’s signal is “for triggering activation or deactivation of one or more AI and/or ML models…” and after the UE receives this signal, the UE transmits “a second signal comprising information associated with activation or deactivation of the one or more AI and/or ML models’ – i.e. model status information, not ML data” and Garcia Rodriguez does not disclose an indication to activate transmission of ML data” and that Li’s “reports are output of ML execution, not ML data input”. The Examiner respectfully disagrees.
Li’s updated machine learning parameters are outputs of the UE local training and are reported so NG-RAN can update the centralized model, making them inputs to that update (see par [0033] and [0051] of Li). Also, Garcia Rodriguez’s second signal is not mapped as the claimed ML data as argued by applicant. Garcia Rodriguez, in the example of figure 3 signaling with par [0089]-[0091], discloses the base station 210 sends UE 220 a request/suggestion of a configuration (e.g., the activation/deactivation) of one or more specific AI/ML models for communication, and further discloses messaging indicating the UE to identify and report back information on the best performing model (par [0172]); thus the signaling of Garcia Rodriguez directs/indicates both model activation and transmission of model information/data to the base station. The combination of Li and Garcia Rodriguez does not read the model activation as data transmission; it uses the activation request to control Li’s configured parameter reporting. Therefore, the combination of Li and Garcia Rodriguez discloses the claimed limitation in question above and the arguments are not persuasive. The rejection is maintained, see detailed rejection for claims 1 and 12 below with more clarifications.
Regarding the dependent claims, which applicant argues based on respective independent claims arguments above, please see responses above.
Response to Amendments
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 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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied 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.
Claim(s) 1, 2, 4, 5, 8, 9, 12, 14, 15 and 18-22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. (US 20240349082 A1) in view of Ly et al. (US 20250310214 A1).
Consider claim 1, Li discloses a user equipment (UE) configured for wireless communications (read as the communication between UE device and network device, figure 1 and 4, par [0067] and [0121]), comprising:
one or more memories comprising instructions; and one or more processors, individually or collectively, configured to execute the instructions to cause the network entity to (read as one or more processors and memory/storage devices storing instruction that causes circuitry associated to perform one or more methodologies, individually or collectively, figures 5, par [0132] and [0136]-[0137]):
receive a configuration for at least one machine learning function name (MLFN) (read as the UE receives ML configuration 114 for a requested machine learning service type, and the service type identifies the ML service or function for which the ML model and configuration are provided, which corresponds to claimed MLFN, figure 1, par [0029], [0050] and [0067]);
transmit an indication to a network entity to request transmission of machine learning (ML) data associated with the at least one MLFN (read as the UE transmitting MachineLearningModelUpdateRequest to the network to request a parameter update for the corresponding model, after which the network sends MachineLearningModelUpdateResponse containing MachineLearningModelUpdate and the corresponding service type; the request thus indicates that the UE requests transmission of the service associated ML data, which corresponds to the claimed ML data associated with the at least one MLFN, figure 1, par [0060] and [0068]);
receive the ML data associated with the at least one MLFN from the network entity (read as the UE receiving MachineLearningModelUpdate containing network generated hidden layer, weight and gradient information; the response includes the corresponding service type, associating the ML parameter data with the identified ML service, par [0004] and [0060]); and
use the ML data as an input for at least one of: operation or training of an ML model associated with the at least one MLFN (read as using the received Machine Learning Model Parameter Updates, including hidden layer, weight and gradient information, for machine learning model update and iteration during local model training, par [0044] and [0060]).
However, Li discloses the claimed invention above and UE’s transmitted update request indicates that the UE requests transmission of the service associated ML data (MachineLearningModelUpdate and the corresponding service type) (figure 1, par [0060] and [0068]) and receive the update response including the corresponding ML service type, associating the ML parameter data with the identified ML service (par [0004] and [0060]) and permit model and parameter updates over a PDU session (see par [0020]) but does not specifically disclose the UE indication having the function of activating transmission of ML data associated with the at least one MLFN (i.e. service associated).
Nonetheless, Ly discloses a UE sending a PDU session establishment request containing an indication that the PDU session is for transmitting or receiving AI or ML traffic; following acceptance in response to the request, the UE communicates AI or ML traffic over the session according to AI or ML policy information, and the traffic includes model training results and training data; thus, the request with the indication initiates establishment of the communication path used for the AI or ML traffic including model training results and training data, figure 4, par [0093] and [0228].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Ly into the teachings of Li, to modify Li’s model update/response exchange between the UE and network to include UE requested PDU session establishment using Ly’s PDU session establishment request containing an AI or ML traffic indication, with the network device configured to initiate transmission of the requested parameter update to the UE device in response to the update request over the established session, in order to deliver Li’s requested service associated parameter update (i.e. service associated ML data) over an accepted session designated for AI or ML traffic communication (see par [0228] of Ly).
Consider claim 2, as applied to claim 1 above, Li, as modified by Ly, discloses wherein: the operation of the ML model comprises at least one of: running the ML model, switching the ML model, or monitoring the ML model; and the training of the ML model comprises: labeling the ML data for a particular UE and network entity setting or configuration, and performing the training of the ML model using the labeled ML data (read as once the UE device 104 has executed the ML model and generated corresponding outputs, the UE device 104 may generate and send a ML report 118 to the network device 102; the ML report 118 may indicate predictions prior to the ML execution 116, outcomes of the ML execution 116 (e.g., actual outputs of the ML execution 116), and requested or selected actions (e.g., an action space) for the UE device 104 to perform based on the ML execution 116; the ML model and/or configuration may be trained and/or updated by the network device 102 and/or the UE device 104, figures 1 and 2, par [0067]-[0068])
Consider claim 4, as applied to claim 1 above, Li, as modified by Ly, discloses wherein: the ML data is applicable for one or more other UEs in a network entity coverage or target area (read as the NG-RAN (e.g., a gNB device) may generate and send a ML capability indication to UEs to indicate that the NG-RAN supports ML (e.g., and may facilitate ML operations at the UE), par [0017]), and the ML data indicates at least one of: an ML model identification (ID) or a model structure (MS) ID together with the ML data (read as the ML models can also be in the form of identifiers for UE to download the models; according to different UE capability, NG-RAN may assign models with smaller granularity, par [0042]).
Consider claim 5, as applied to claim 4 above, Li, as modified by Ly, discloses wherein the ML data is received via at least one of: a system information block (SIB) or multicast broadcast service (MBS) over MBS channel (read as the network's ML capability may be indicated by a new SIB, this new SIB (e.g. SIBX) contains information related to machine learning, par [0060]-[0062]).
Consider claim 8, as applied to claim 1 above, Li, as modified by Ly, discloses wherein one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: transmit a request to a network entity for the ML data, wherein the ML data is received by the UE in response to the request (read as identify a service registration, received from the UE device, indicating that the UE device requests machine learning support from the node B device; transmit, to the UE device, a request for information associated with the UE device, the information associated with at least one of hardware capabilities or machine learning capabilities of the UE device; identify the information received from the UE based on the request for information, see abstract par [0067]-[0068]).
Consider claim 9, as applied to claim 8 above, Li, as modified by Ly, discloses wherein the transmit comprises transmit the request via at least one of: UE assistance information (UAI) or a subscribe request (read as identify a service registration, received from the UE device, indicating that the UE device requests (i.e. subscribe request as UE is subscribed to base station) machine learning support from the node B device, see abstract par [0067]-[0068]).
Consider claim 10, as applied to claim 8 above, Li, as modified by Ly, discloses wherein the request indicates at least one of: the at least one MLFN; an ML model identification (ID); a model structure (MS) ID; geographical information comprising at least one of: current geographical area of the UE, a public land mobile network (PLMN), cell information, or a frequency list; a validity time comprising at least one of: a duration time or an interval time at which the ML data has to be provided to the UE; one or more network configurations or settings; or a type of the ML data comprising at least one of: meta data, training data, or inference data (read as UE(s) send their interested service type/ID to the network and requesting such service(s), par [0063]).
Consider claim 12, Li discloses a network entity configured for wireless communications (read as the communication between UE device and network device, figure 1 and 4, par [0067] and [0121]), comprising:
one or more memories comprising instructions; and one or more processors, individually or collectively, configured to execute the instructions to cause the network entity to (read as one or more processors and memory/storage devices storing instruction that causes circuitry associated to perform one or more methodologies, individually or collectively, figures 5, par [0132] and [0136]-[0137]):
transmit a configuration for at least one machine learning function name (MLFN) to at least one user equipment (UE) (read as the UE receives ML configuration 114 transmitted by the network device for a requested machine learning service type, and the service type identifies the ML service or function for which the ML model and configuration are provided, which corresponds to claimed MLFN, figure 1, par [0029], [0050] and [0067]);
determine machine learning (ML) data associated with the at least one MLFN (read as the network device determine the MachineLearningModelUpdate and send it to the UE, the UE receiving MachineLearningModelUpdate containing network generated hidden layer, weight and gradient information; the response includes the corresponding service type, associating the ML parameter data with the identified ML service, par [0004] and [0060]) to be used as an input for at least one of: operation or training of an ML model associated with the at least one MLFN (read as using the received Machine Learning Model Parameter Updates, including hidden layer, weight and gradient information, for machine learning model update and iteration during local model training, par [0044] and [0060]);
receive an indication from the at least one UE to request transmission of the ML data to the at least one UE (read as the UE transmitting MachineLearningModelUpdateRequest to the network to request a parameter update for the corresponding model, after which the network sends MachineLearningModelUpdateResponse containing MachineLearningModelUpdate and the corresponding service type; the request thus indicates that the UE requests transmission of the service associated ML data, which corresponds to the claimed ML data associated with the at least one MLFN, figure 1, par [0060] and [0068]); and
transmit the ML data to the at least one UE (read as the network device determine the MachineLearningModelUpdate and send it to the UE, the UE receiving MachineLearningModelUpdate containing network generated hidden layer, weight and gradient information; the response includes the corresponding service type, associating the ML parameter data with the identified ML service, par [0004] and [0060]).
However, Li discloses the claimed invention above and UE’s transmitted update request indicates that the UE requests transmission of the service associated ML data (MachineLearningModelUpdate and the corresponding service type) (figure 1, par [0060] and [0068]) and receive the update response including the corresponding ML service type, associating the ML parameter data with the identified ML service (par [0004] and [0060]) and permit model and parameter updates over a PDU session (see par [0020]) but does not specifically disclose the UE indication having the function of activating transmission of ML data associated with the at least one MLFN (i.e. service associated).
Nonetheless, Ly discloses a UE sending a PDU session establishment request containing an indication that the PDU session is for transmitting or receiving AI or ML traffic; following acceptance in response to the request, the UE communicates AI or ML traffic over the session according to AI or ML policy information, and the traffic includes model training results and training data; thus, the request with the indication initiates establishment of the communication path used for the AI or ML traffic including model training results and training data, figure 4, par [0093] and [0228].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Ly into the teachings of Li, to modify Li’s model update/response exchange between the UE and network to include UE requested PDU session establishment using Ly’s PDU session establishment request containing an AI or ML traffic indication, with the network device configured to initiate transmission of the requested parameter update to the UE device in response to the update request over the established session, in order to deliver Li’s requested service associated parameter update (i.e. service associated ML data) over an accepted session designated for AI or ML traffic communication (see par [0228] of Ly).
Consider claim 14, as applied to claim 12 above, Li, as modified by Ly, discloses wherein: the at least one UE corresponds to a first UE and a second UE; and the first UE and the second UE requiring same ML data inputs or controls (read as the NG-RAN (e.g., a gNB device) may generate and send a ML capability indication to UEs (i.e. first UE and second UE) to indicate that the NG-RAN supports ML (e.g., and may facilitate ML operations at the UE), par [0017]) associated with at least one of: the at least one MLFN, an ML model identification (ID), or a model structure (MS) ID (read as the ML models can also be in the form of identifiers for UE to download the models; according to different UE capability, NG-RAN may assign models with smaller granularity, par [0042]).
Consider claim 15, as applied to claim 14 above, Li, as modified by Ly, discloses wherein the ML data is transmitted to the first UE and the second UE via at least one of: a system information block (SIB) or a multicast broadcast service (MBS) (read as the network's ML capability may be indicated by a new SIB, this new SIB (e.g. SIBX) contains information related to machine learning, par [0060]-[0062]).
Consider claim 18, as applied to claim 12 above, Li, as modified by Ly, discloses wherein: the at least one UE corresponds to a first UE; the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: receive a request for the ML data from the first UE, wherein the ML data is transmitted to the first UE in response to the request (read as identify a service registration, received from the UE device, indicating that the UE device requests machine learning support from the node B device; transmit, to the UE device, a request for information associated with the UE device, the information associated with at least one of hardware capabilities or machine learning capabilities of the UE device; identify the information received from the UE based on the request for information, see abstract par [0067]-[0068]).
Consider claim 19, as applied to claim 18 above, Li, as modified by Ly, discloses wherein the request is received via at least one of: UE assistance information (UAI) or a subscribe request (read as identify a service registration, received from the UE device, indicating that the UE device requests (i.e. subscribe request as UE is subscribed to base station) machine learning support from the node B device, see abstract par [0067]-[0068]).
Consider claim 20, as applied to claim 18 above, Li, as modified by Ly, discloses wherein the request indicates at least one of: the at least one MLFN; an ML model identification (ID); a model structure (MS) ID; geographical information comprising at least one of: current geographical area of the first UE, a public land mobile network (PLMN), cell information, or a frequency list; a validity time comprising at least one of: a duration time or an interval time at which the ML data has to be provided to the first UE; one or more network configurations; or a type of the ML data comprising at least one of: meta data, training data, or inference data (read as UE(s) send their interested service type/ID to the network and requesting such service(s), par [0063]).
Consider claim 21, as applied to claim 18 above, Li, as modified by Ly, discloses wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: receive an indication from the first UE to activate or deactivate transmission of the ML data to the first UE (read as the UE sends UE capability information 112 to base station 102 to cause/activate ML configuration 114 from the base station to UE, par [0067]-[0068]).
Consider claim 22, as applied to claim 18 above, Li, as modified by Ly, discloses wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: transmit the request to another network entity when one or more conditions are satisfied (read as once RAN accepts the handover request, NG-RAN will send a granted response to the request UE and forward the corresponding UE context (i.e. the UE request) to the target cell, par [0021]).
Claim(s) 24, 25, 27, 28 and 30 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. (US 20240349082 A1) in view of Garcia Rodriguez et al. (US 20250203401 A1).
Consider claim 24, Li discloses a user equipment (UE) configured for wireless communications (read as the communication between UE device and network device, figure 1 and 4, par [0067] and [0121]), comprising:
one or more memories comprising instructions; and one or more processors, individually or collectively, configured to execute the instructions to cause the network entity to (read as one or more processors and memory/storage devices storing instruction that causes circuitry associated to perform one or more methodologies, individually or collectively, figures 5, par [0132] and [0136]-[0137]):
transmit UE capability information to a network entity (read as UE device 104 transmitting UE capability information 112, including ML or hardware capabilities, to network device 102, figure 1, par [0067]);
receive from the network entity a configuration for at least one machine learning function name (MLFN) and a request for machine learning (ML) data associated with the at least one MLFN to be used as an input for at least one of: operation or training of an ML model associated with the at least one MLFN, wherein the configuration and the request are based on the UE capability information (read as the network device 102 selects and sends ML configuration 114 based on UE capability information 112 and the requested ML service, the UE 102 then receives the ML configuration 114, the service type included in the configuration identifies the ML service or function and corresponds to the MLFN; the capability selected MachineLearningConfiguration contains MachineLearningReportConfiguration, which specifies the report type and periodicity; configuring ML model parameter updates on the UE based on the configuration functionally requests the UE to report the corresponding service associated parameters produced through local training so that NG-RAN can update the centralized model (i.e. report as input for the update); since the reporting is part of the ML configuration 114 selected based on UE capability information 112, both the configuration and reporting request are based on UE capability information 112, figure 1, par [0029], [0031], [0033], [0049], [0051] and [0067]);
receive an indication from the network entity to configure transmission of the ML data associated with the at least one MLFN (read as the UE receives a Machine Learning Report Configuration from NG-RAN that indicates the type, frequency and start offset of its model parameter report; the report configuration thus reasonably functions are an indication configuring transmission of ML data associated with the registered service type, par [0031], [0035] and [0051]); and
transmit the ML data to the network entity (read as the UE transmits a MachineLearningReport in the UE to network direction containing the corresponding locally trained model parameter updates, which NG-RAN uses to update the centralized model, par [0033], [0051], [0057 and [0059]).
However, Li discloses the claimed invention above with the UE receiving Machine Learning Report Configuration specifying a model parameter report and its timing, and then transmission/reporting of locally trained parameters for the service (i.e. input) so that the NG-RAN can update the centralized model (par [0031], [0033], [0035] and [0051]) but does not specifically disclose a separate indication from the network entity to activate transmission of the ML data associated with the at least one MLFN.
Nonetheless, Garcia Rodriguez discloses the UE 112 receiving a first signal from network node 110 that triggers mode activation and transmitting a second signal containing information associated with that activation; the triggering information may identify the model, specify activation conditions and a model retraining purpose and indicate whether a response is expected (figure 18, par [0116], [0317], [0334] and [0336]); for example, in the figure 3 signaling with par [0089]-[0091], the base station 210 sends UE 220 a request/suggestion of a configuration (e.g., the activation/deactivation) of one or more specific AI/ML models for communication and/or performance evaluation purposes, the 220 performs the requested actions and sends an activation notification to the base station 210; and further discloses message indicating the UE to identify and report back information on the best performing model (par [0172]); thus the signaling direct/indicate both model activation and transmission of model information/data to the base station.
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Garcia Rodriguez into the teachings of Li, to modify Li’s model parameter reporting process using Garcia Rodriguez’s model activation signaling and indication to report back model information to trigger transmission of configured parameter report, in order to coordinate UE parameters (ML data) reporting with requested model evaluation/retraining so that NG-RAN receives the resulting parameters for centralized model updating (see par [0172] and [0317] of Garcia Rodriguez).
Consider claim 25, as applied to claim 24 above, Li, as modified by Garcia Rodriguez, discloses wherein the configuration further indicates at least one of: an ML model identification (ID) or a model structure (MS) ID (read as the ML models can also be in the form of identifiers for UE to download the models; according to different UE capability, NG-RAN may assign models with smaller granularity, par [0042]).
Consider claim 27, Li discloses a network entity configured for wireless communications (read as the communication between UE device and network device, figure 1 and 4, par [0067] and [0121]), comprising:
one or more memories comprising instructions; and one or more processors, individually or collectively, configured to execute the instructions to cause the network entity to (read as one or more processors and memory/storage devices storing instruction that causes circuitry associated to perform one or more methodologies, individually or collectively, figures 5, par [0132] and [0136]-[0137]):
receive user equipment (UE) capability information from a UE (read as UE device 104 transmitting UE capability information 112 to the network device to receive, including ML or hardware capabilities, to network device 102, figure 1, par [0067]);
transmit to the UE a configuration for at least one machine learning function name (MLFN) and a request for machine learning (ML) data associated with the at least one MLFN to be used as an input for at least one of: operation or training of an ML model associated with the at least one MLFN, wherein the configuration and the request are based on the UE capability information (read as the network device 102 selects and sends ML configuration 114 based on UE capability information 112 and the requested ML service, the UE 102 then receives the ML configuration 114, the service type included in the configuration identifies the ML service or function and corresponds to the MLFN; the capability selected MachineLearningConfiguration contains MachineLearningReportConfiguration, which specifies the report type and periodicity; configuring ML model parameter updates on the UE based on the configuration functionally requests the UE to report the corresponding service associated parameters produced through local training so that NG-RAN can update the centralized model (i.e. report as input for the update); since the reporting is part of the ML configuration 114 selected based on UE capability information 112, both the configuration and reporting request are based on UE capability information 112, figure 1, par [0029], [0031], [0033], [0049], [0051] and [0067]);
transmit an indication to the UE to configure transmission of the ML data associated with the at least one MLFN (read as the UE receives a Machine Learning Report Configuration from NG-RAN that indicates the type, frequency and start offset of its model parameter report; the report configuration thus reasonably functions are an indication configuring transmission of ML data associated with the registered service type, par [0031], [0035] and [0051]); and
receive the ML data from the UE (read as the UE transmits a MachineLearningReport in the UE to network direction containing the corresponding locally trained model parameter updates, which NG-RAN uses to update the centralized model, par [0033], [0051], [0057 and [0059]).
However, Li discloses the claimed invention above with the UE receiving Machine Learning Report Configuration specifying a model parameter report and its timing, and then transmission/reporting of locally trained parameters for the service (i.e. input) so that the NG-RAN can update the centralized model (par [0031], [0033], [0035] and [0051]) but does not specifically disclose a separate indication from the network entity to activate transmission of the ML data associated with the at least one MLFN.
Nonetheless, Garcia Rodriguez discloses the UE 112 receiving a first signal from network node 110 that triggers mode activation and transmitting a second signal containing information associated with that activation; the triggering information may identify the model, specify activation conditions and a model retraining purpose and indicate whether a response is expected (figure 18, par [0116], [0317], [0334] and [0336]); for example, in the figure 3 signaling with par [0089]-[0091], the base station 210 sends UE 220 a request/suggestion of a configuration (e.g., the activation/deactivation) of one or more specific AI/ML models for communication and/or performance evaluation purposes, the 220 performs the requested actions and sends an activation notification to the base station 210; and further discloses message indicating the UE to identify and report back information on the best performing model (par [0172]); thus the signaling direct/indicate both model activation and transmission of model information/data to the base station.
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Garcia Rodriguez into the teachings of Li, to modify Li’s model parameter reporting process using Garcia Rodriguez’s model activation signaling and indication to report back model information to trigger transmission of configured parameter report, in order to coordinate UE parameters (ML data) reporting with requested model evaluation/retraining so that NG-RAN receives the resulting parameters for centralized model updating (see par [0172] and [0317] of Garcia Rodriguez).
Consider claim 28, as applied to claim 27 above, Li, as modified by Garcia Rodriguez, discloses wherein the configuration further indicates at least one of: an ML model identification (ID) or a model structure (MS) ID (read as the ML models can also be in the form of identifiers for UE to download the models; according to different UE capability, NG-RAN may assign models with smaller granularity, par [0042]).
Consider claim 30, as applied to claim 27 above, Li, as modified by Garcia Rodriguez, discloses wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: determine the ML data to be reported based on the UE capability information (read as based on the UE capability information 112, the network device 102 may select or generate a ML configuration 114 for the UE device 104, and may send the ML configuration 114 to the UE device; for example, based on any requested service in the service registration 108, the network device 102 may select a ML model and a ML configuration for the ML model for the service based on the UE capability information 112 (e.g., a ML model/configuration using more or fewer resources depending on the UE capability information 112), par [0067]).
Claims 3 and 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. (US 20240349082 A1) in view of Ly et al. (US 20250310214 A1), and in further view of Wang et al. (US 20230325679 A1).
Consider claim 3, as applied to claim 1 above, Li, as modified by Ly, discloses wherein: the ML data comprises model management data and training data (read as the device may generate/select and transmit, to the UE device, a machine learning configuration and a ML model (e.g., the ML configuration 114 of FIG. 1) based on the information received from the UE device; and the ML model and/or configuration may be trained and/or updated, par [0068] and [0080]) and the training data indicates at least one of: a network load, an event threshold, or layer 1 (L1) and layer 2 (L2) measurements (read as, for example, the reportPeriodicity (Value ms20), see Table 2, par [0050]; or required memory size (e.g., unit: MB/KB) as described in par ]0062]) but does not specifically disclose the model management data indicates at least one of: a number of antennas or a number of retransmissions.
Nonetheless, in related art, Wang discloses a machine learning system between base station and different UEs, wherein the UEs configures/indicates its number of antennas based on ML configuration parameters from base station 120, figure 6, par [0093]-[0094].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Wang into the teachings of Li, which modified by Ly, to design the base station to configure the number of UE antennas in order to accurately configure/control the UE for ML model training.
Consider claim 13, as applied to claim 12 above, Li, as modified by Ly, discloses wherein: the ML data comprises model management data and training data (read as the device may generate/select and transmit, to the UE device, a machine learning configuration and a ML model (e.g., the ML configuration 114 of FIG. 1) based on the information received from the UE device; and the ML model and/or configuration may be trained and/or updated, par [0068] and [0080]) and the training data indicates at least one of: a network load, an event threshold, or layer 1 (L1) and layer 2 (L2) measurements (read as, for example, the reportPeriodicity (Value ms20), see Table 2, par [0050]; or required memory size (e.g., unit: MB/KB) as described in par ]0062]) but does not specifically disclose the model management data indicates at least one of: a number of antennas or a number of retransmissions.
Nonetheless, in related art, Wang discloses a machine learning system between base station and different UEs, wherein the UEs configures/indicates its number of antennas based on ML configuration parameters from base station 120, figure 6, par [0093]-[0094].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Wang into the teachings of Li, which modified by Ly, to design the base station to configure the number of UE antennas in order to accurately configure/control the UE for ML model training.
Claims 6, 7, 10, 11, 16 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. (US 20240349082 A1) in view of Ly et al. (US 20250310214 A1), and in further view of Zeng et al. (US 20230209390 A1).
Consider claim 6, as applied to claim 1 above, Li, as modified by Ly, discloses the claimed invention above but does not specifically disclose wherein the ML data is received via a unicast message when at least one condition is satisfied: the UE connects to a network entity; or the UE is configured with at least one of: the at least one MLFN, an ML model identification (ID), or a model structure (MS) ID.
Nonetheless, in related art, Zeng discloses a UE connected the base station, and the base station would send the artificial intelligence (AI) task to the UE in a unicast form, par [0008]) and abstract.
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Ly, to design the base station to send the ML configuration (AI task) to the UE for ML training as unicast would ensure reliable one-to-one communication.
Consider claim 7, as applied to claim 6 above, Li, as modified by Ly, discloses the claimed invention above the UE sends UE capability information 112 to base station 102 to cause/activate ML configuration 114 from the base station to UE (par [0067]-[0068]) but does not specifically disclose wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: transmit an indication using a medium access control (MAC) control element (CE), UE assistance information (UAI), or uplink control information (UCI) to the network entity to activate or deactivate transmission of the ML data to the UE.
Nonetheless, in related art, Zeng discloses the UE reports the UE capability to the network device and UE capability is carried over the MAC CE message, par [0211].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Ly, to design the UE to send the UE capability to the base station over MAC CE message as MAC CE offers low latency.
Consider claim 10, as applied to claim 8 above, Li, as modified by Ly, discloses wherein the request indicates at least one of: the at least one MLFN; an ML model identification (ID); a model structure (MS) ID; geographical information comprising at least one of: current geographical area of the UE, a public land mobile network (PLMN), cell information, or a frequency list; a validity time comprising at least one of: a duration time or an interval time at which the ML data has to be provided to the UE; one or more network configurations or settings; or a type of the ML data comprising at least one of: meta data, training data, or inference data (read as the machine learning service type/ID send by UE as a request to the base station, par [0063]).
Consider claim 11, as applied to claim 1 above, Li, as modified by Ly, discloses the claimed invention above the UE sends UE capability information 112 to base station 102 to cause/activate ML configuration 114 from the base station to UE (par [0067]-[0068]) but does not specifically disclose wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: transmit an indication using a medium access control (MAC) control element (CE), UE assistance information (UAI), or uplink control information (UCI) to the network entity to activate or deactivate transmission of the ML data to the UE.
Nonetheless, in related art, Zeng discloses the UE reports the UE capability to the network device and UE capability is carried over the MAC CE message, par [0211].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Ly, to design the UE to send the UE capability to the base station over MAC CE message as MAC CE offers low latency.
Consider claim 16, as applied to claim 12 above, Li, as modified by Ly, discloses claimed invention above and wherein: the at least one UE corresponds to a first UE (read as the UE , par [0067]-[0068] but does not specifically disclose the ML data is transmitted to the first UE via a unicast message.
Nonetheless, in related art, Zeng discloses a UE connected the base station, and the base station would send the artificial intelligence (AI) task to the UE in a unicast form, par [0008]) and abstract.
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Ly, to design the base station to send the ML configuration (AI task) to the UE for ML training as unicast would ensure reliable one-to-one communication.
Consider claim 17, as applied to claim 16 above, Li, as modified by Ly, discloses the claimed invention above the UE sends UE capability information 112 to base station 102 to cause/activate ML configuration 114 from the base station to UE (par [0067]-[0068]) but does not specifically disclose wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: transmit an indication using a medium access control (MAC) control element (CE), UE assistance information (UAI), or uplink control information (UCI) to the network entity to activate or deactivate transmission of the ML data to the UE.
Nonetheless, in related art, Zeng discloses the UE reports the UE capability to the network device and UE capability is carried over the MAC CE message, par [0211].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Ly, to design the UE to send the UE capability to the base station over MAC CE message as MAC CE offers low latency.
Claims 26 and 29 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. (US 20240349082 A1) in view of Garcia Rodriguez et al. (US 20250203401 A1), and in further view of Zeng et al. (US 20230209390 A1).
Consider claim 26, as applied to claim 24 above, Li, as modified by Garcia Rodriguez, discloses the claimed invention above and an indication from the network entity to activate or deactivate transmission of the ML data to the network entity (read as the network node 110 transmitting to UE 112 a first signal for triggering activation or deactivation of one or more AI and/or ML models by the UE 112, the UE receiving that first signal and then transmitting a second signal, and the triggering information including activation =information, par [0116], [0317], [0334] and [0336] of Garcia Rodriguez) but does not specifically disclose wherein the processor is further configured to execute the computer-executable instructions and cause the UE to: receive, via a medium access control (MAC) control element (CE) or a downlink control information (DCI), the indication.
Nonetheless, in related art, Zeng discloses the base station and the UE are using MAC CE to indicate/report collected data to each other, par [0343].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Garcia Rodriguez, as modified by Ly, to design the network device to send ML capability indication 106 to the UE device over MAC CE as MAC CE offers low latency.
Consider claim 29, as applied to claim 27 above, Li, as modified by Garcia Rodriguez, discloses the claimed invention above and wherein the one or more processors, individually or collectively, configured to execute the instructions to cause the UE to: transmit an indication to the UE to activate or deactivate transmission of the ML data to the network entity (read as the network node 110 transmitting to UE 112 a first signal for triggering activation or deactivation of one or more AI and/or ML models by the UE 112, the UE receiving that first signal and then transmitting a second signal, and the triggering information including activation =information, par [0116], [0317], [0334] and [0336] of Garcia Rodriguez) but does not specifically disclose transmit the indication via a medium access control (MAC) control element (CE) or a downlink control information (DCI).
Nonetheless, in related art, Zeng discloses the base station and the UE are using MAC CE to indicate/report collected data to each other, par [0343].
Therefore, it would have been obvious for a person with ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Zeng into the teachings of Li, as modified by Garcia Rodriguez, to design the network device to send ML capability indication 106 to the UE device over MAC CE as MAC CE offers low latency.
Allowable Subject Matter
Claim 23 is objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
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 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Junpeng Chen whose telephone number is (571) 270-1112. The examiner can normally be reached on Monday - Thursday, 8:00 a.m. - 5:00 p.m., 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, Anthony S Addy can be reached on 571-272-7795. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/Junpeng Chen/
Primary Examiner, Art Unit 2645
Prosecution Insights