DETAILED ACTION
1. This action is responsive to the communications filed on 05/26/2026.
2. Claims 21-35 are pending in this application.
3. Claims 21-31, 33-35 have been amended.
4. Claims 1-20 have been previously cancelled.
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
5. Applicant’s arguments with respect to claim 21 have been considered but are moot in view of the new grounds of rejection.
Although a new ground of rejection has been used to address additional limitations that have been added to claim 21, a response is considered necessary for several of applicant's arguments since reference Ren will continue to be used to meet several claimed limitations.
In the remarks, applicant argued that:
a. Additionally, the Office Action maps Ren's model update to the claimed support frame and the claimed parameters to Ren's parameters (see Office Action at p. 5 and Ren at paragraphs 0060 and 0061). However, Ren's model update uses the parameters to generate the model update (see Ren at paragraph 0061 stating that "[i]n generating the model update, local model updating component 242 can include parameters corresponding to the updated FL model"). Ren's generation using the model updates is done at step 404, while transmission of the model update is done at step 408 (see Ren at FIG. 4 and paragraphs 0060-0063). Therefore any model update based on the set of parameters would be done before transmission of the model update (e.g., mapped to the support frame) in Ren. It is respectfully requested that the rejections be withdrawn (Applicant’s remarks, page 7).
In response: The examiner respectfully disagrees.
In Ren, the UE will generate a model update to be applied based on the local training (Figure 4, 404). The UE then transmits the update to the base station (Figure 4, 408). When the base station receives the model update (Figure 5, 502), it then generates a converged model update that includes the model update sent by the UE(s) (Figure 5, 504). This converged model update is the claimed ‘model update’ and is clearly done after transmission of the model updates from each UE.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
6. Claims 21-35, are rejected under 35 U.S.C. 103 as being unpatentable over Ren et al (US 2024/0256898) in view of Pezeshki et al. (US 2023/0080218).
Regarding claim 21, Ren disclosed:
A method performed by a station (Figure 2, UE 104), the method comprising:
receiving, via an access point (AP) (Paragraph 41, access point) associated with a basic service set (BSS) (Paragraph 41, basic service set) in a wireless local area network (WLAN) (Paragraph 36, WiFi), a learning model announcement frame (Paragraph 59, indication of FL model) indicating a model sharing process and soliciting a support frame (Paragraph 60, model update) (Paragraph 41, the UE may be referred to as station. Base station that is connected to UE includes a basic service set along with an access point. Paragraph 59, an indication of an FL model is received from a base station. The base FL is transmitted to various UEs so that the various UEs can perform a local model updating process (i.e., model sharing process). Paragraph 60, a model update is generated based on the local training after receiving an indication of the FL model from the base station (i.e., soliciting));
transmitting (Paragraph 63, transmitting), in response to the learning model announcement frame, a support frame (Paragraph 60, model update) indicating participation in the model sharing process (Paragraph 60, local training to update the model), the support frame comprising a set of parameters (Paragraph 61, parameters) associated with a model (Paragraph 60, Figure 4, a model update to be applied to the FL model is generated based on local training on the FL model after receiving the indication of the FL model from the base station. The local model updating component 242 (of the UE) generates the model update to be applied to the FL model. The updates from the various UEs are collected and applied to the base model. Paragraph 63, a report of the model update is transmitted).
While Ren disclosed sending a model update to be applied to the FL model (see above), Ren did not explicitly disclose wherein the model is stored at the station before receiving the learning model announcement frame; and after transmitting the support frame, updating the model based at least upon the participation in the sharing model process and the set of parameters associated with the model.
However, in an analogous art, Pezeshki disclosed wherein the model is stored at the station before receiving the learning model announcement frame (Paragraph 61, the UE locally trains the machine learning component using training data collected by the UEs. The UE trains the machine learning component by optimizing a set of model parameters, w(n) where n is the federated learning round index (i.e., continually using the model for a series of rounds). Paragraph 62, the initial provisioning of a machine learning component on a UE or the transmission of a global update to the machine learning component to a UE triggers the beginning of a new round of federated learning. Paragraph 63, the UEs collect training data and store it in memory devices. The stored training data is referred to as the local dataset. Paragraph 72, the UE receives a machine learning component from the base station for the second time (see Figure 3). The machine learning component includes a federated learning configuration which configures a participation indication to be used by the UE to indicate a participation status of the UE associated with at least one federated learning round. The participation indication indicates whether the UE participated in a prior federated learning round. Therefore, once the base station transmits the machine learning component the first time, the UE stores it locally for subsequent training and update rounds. The machine learning component is equated to the model as it is what is being trained using federated learning (see paragraphs 56-57));
after transmitting the support frame, updating the model based at least upon the participation in the sharing model process and the set of parameters associated with the model (Paragraph 72, the UE may transmit a participation indication (i.e., support frame) that indicates whether the UE participated in prior federated learning rounds and whether it will participate in future rounds. Once the UE confirms participation (i.e., participation in the sharing model process), the base station sends the machine learning component 315. The UE utilizes the machine learning component to generate a local update 340 (i.e., updating the model), and sends this update back to the base station). Figure 3, further showing that after the local update 1 340 is received, the base station creates another machine learning component 315 that is sent to the UE, where the UE then sends another local update k 340 back to the base station. This shows the iterative process being performed multiple times).
One of ordinary skill in the art would have been motivated to combine the teachings of Ren with Pezeshki because the references involve federated learning, and as such, are within the same environment.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the storing and updating the model of Pezeshki with the teachings of Ren in order to facilitate efficient updating of machine learning components (Pezeshki, Paragraph 71).
Regarding claim 28, the claim is substantially similar to claim 21. Claim 28 recites a transceiver and a processor (Ren, Figure 2, transceiver 202 and processor 212). Therefore, the claim is rejected under the same rationale.
Regarding claims 22, 29, the limitations of claims 21, 28, have been addressed. Ren and Pezeshki disclosed:
wherein the model comprises at least one of a federated learning (FL) model or a machine learning (ML) model (Ren, Paragraph 59, the FL model can be a ML model).
Regarding claims 23, 30, the limitations of claims 21, 28, have been addressed. Ren and Pezeshki disclosed:
wherein the model comprises a first model, and wherein the method comprises: receiving a second support frame from a second station, the second support frame comprising a second set of parameters associated with a second model (Ren, Paragraph 75, Figure 7, showing multiple UEs that relay model updates to the relay UE); and
updating the first model based on the second set of parameters (Ren, Paragraph 75, converging the local model updates into a relay model update sent to the base station).
Regarding claims 24, 31, the limitations of claims 21, 28, have been addressed. Ren and Pezeshki disclosed:
wherein the model announcement frame comprises announcement parameters, the announcement parameters comprising at least one of a model identifier (ID) (Ren, Paragraph 65, model identifier), a number of model layers, or a number of weights per layer, the method further comprising configuring the station to update the model in accordance with the announcement parameters (Ren, Paragraph 65, the model update includes reports with model parameters and corresponding values. Paragraph 66, a converged model update is generated based on the parameters within the model update).
Regarding claims 25, 33, the limitations of claims 21, 28, have been addressed. Ren and Pezeshki disclosed:
further comprising: receiving a gradient update for the model (Pezeshki, Paragraphs 66-68, using a stochastic gradient descent (SGD) algorithm that will determine the gradients. UE will transmit a compressed set of gradients of their local updates); and
configuring the station to update the model in accordance with the received gradient update (Pezeshki, Paragraph 68, the base station will aggregate the updates from the UE and average the gradients to determine an aggregated update. Paragraph 70, updating the global machine learning component with the gradient updates and transmit an updated associated with the global machine learning component to the UEs).
For motivation, please refer to claim 21.
Regarding claims 26, 34, the limitations of claims 21, 28, have been addressed. Ren and Pezeshki disclosed:
further comprising: training the model; and transmitting results of the training of the model (Ren, Paragraph 60, based on the local training, a model update to be applied to the FL model is generated for the FL model. Paragraph 71, the model update is transmitted to an upstream node).
Regarding claims 27, 35, the limitations of claims 21, 28, have been addressed. Ren and Pezeshki disclosed:
further comprising: receiving a message comprising a target wake time (TWT); and configuring the station to receive training parameters for the model during the TWT (Ren, Paragraph 65, receiving the update reports from multiple UEs at a similar time. This is based on a time (i.e., TWT) or trigger detected by the UEs).
Regarding claim 32, the limitations of claim 28 have been addressed. Ren and Pezeshki disclosed:
wherein the announcement frame comprises at least one of an uplink (UL) schedule or a downlink (DL) schedule, the processor further configured to configure the station to transmit and receive in accordance with at least one of the UL schedule or the DL schedule (Ren, Paragraph 22, federated learning includes downlink signaling to scheduled users. Paragraph 23, UEs are scheduled for FL in downlink (i.e., transmit/receive in accordance with schedule) but some low tier devices may have uplink coverage constraints).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Steven C. Nguyen whose telephone number is (571)270-5663. The examiner can normally be reached M-F 7AM - 3PM and alternatively, through e-mail at Steven.Nguyen2@USPTO.gov.
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, Christopher Parry can be reached at 571-272-8328. 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.
/S.C.N/Examiner, Art Unit 2451
/Chris Parry/Supervisory Patent Examiner, Art Unit 2451
Prosecution Insights