DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
This communication is in response to the amendment filed 12/15/2025. The amendment has been entered and considered.
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.
Claim(s) 1, 4-6, 8, 10, 11, 13-15, 33-35, 37, 39, 40, 42, 43 is/are rejected under 35 U.S.C. 103 as being unpatentable over Jeon et al. “Jeon” US 2025/0016619 in view of Sahin et al. “Sahin” US2022/0391696.
Regarding claims 1, 15, and 43, Jeon teaches a computer readable medium (paragraph 22), method and apparatus for wireless communication at a user equipment (UE) (figure 21), comprising:
memory; and one or more processors coupled to the memory and configured to cause the UE (figure 21) to:
identify, in at least one round of a federated learning procedure, at least one gradient update based on local data and a local copy of a machine learning model associated with the federated learning procedure (Paragraphs 128, 132 and Figure 13 shows instance of a federal learning procedure with gradient updates based on local information); and
transmit, in the at least one round of the federated learning procedure, based on an over-the-air (OTA) aggregation scheme and for a network node, an indication of the at least one gradient update via a set of resources associated with the OTA aggregation scheme, wherein the indication of the at least one gradient update comprises at least one sequence (each device transmits the local parameters used for the gradient update on resources allocated based on control information. The parameters used for the federated learning include weight and gradient of local models; Paragraph 132. The weights can be expressed as a binary sequence; Paragraph 143. The federated learning process utilizes OTA; Paragraph 137.
Jeon does not expressly disclose non-coherent OTA and a transmit power, which is associated with the sequence, is based at least in part on a magnitude of the gradient update. Sahin teaches federated learning for over-the-air aggregated computations using non-coherent detectors (Paragraph; 2, 82 and 84). Sahin further teaches that a weighting function can be used which is for the power of the transmitted OFDM symbols. A gradual power increase is tied directly to the magnitude of the local gradient; Paragraph 66. Everything that occurs in a system is “associated” with everything else, thus when combined with Jeon, the power would be associated with the sequence of Jeon.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include non-coherent OTA systems with transmission power based on the magnitude of the gradient as taught by Sahin.
One would be motivated to make the modification such that the system can ensure gradual power increases as taught by Sahin; Paragraph 66.
Regarding claims 4 and 33, Jeon does not disclose the transmit power is proportional to the magnitude of the at least one gradient update; however, Sahin teaches power is tied to the magnitude of the local gradient; Paragraph 66.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include transmission power is proportional to the magnitude of the gradient as taught by Sahin.
One would be motivated to make the modification such that the system can ensure gradual power increases as taught by Sahin; Paragraph 66.
Regarding claims 5 and 34, Jeon does not disclose the transmit power is based on a sum of the magnitude of the at least one gradient update and an offset which concerts the gradient range to a non-negative range; however, Sahin teaches power is tied to the magnitude of the local gradient. The last equation in Paragraph 66 shows a value “+p” which is being viewed as the offset value. The value of p is a non-negative number thus this would “convert” a gradient range (if negative) to a non-negative range. This equation is with respect to the offset value and magnitude of the gradient which results in power increases. Thus one can see the power is tied/associated/based on the magnitude and offset values as claimed.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include transmission power is proportional to the magnitude of the gradient and offset as taught by Sahin.
One would be motivated to make the modification such that the system can ensure gradual power increases as taught by Sahin; Paragraph 66.
Regarding claims 6 and 35, Jeon teaches the resources have a first and second subset of resources for transmitting the gradient update wherein the sequence is transmit via the first or second resources based on the gradient updating being positive or negative (information is sent using various resources. Configuration information is received by the UE which allocates what resources to use for transmission. Thus one can see there are a plurality of resources and each allocated resource is viewed as a subset (thus at least a first and second set of resources). These resources are used to transmit parameters, weight, gradient, etc.; Paragraph 66. A gradient can only be positive or negative and thus one would see Jeon teaches sending the indication information on a subset of resources (first or second) based on the gradient being positive or negative).
Regarding claims 8 and 37, the Jeon does not disclose the transmission power is based on a local learning weight associated with the UE; however, Sahin teaches the power values are based on a weighted function (i.e. local learning weight); Paragraph 66. With respect to the federal learning model, Sahin teaches local loss function models; Paragraph 56 which is associated with the weighting function of paragraph 66. The Examiner notes, everything that occurs in a system can broadly be assumed to be “based on” everything else that happens in the system. Without defining the exact correlation (or how one impacts the other) there is no real meaning behind the “based on” terminology.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include transmission power is based on a local learning weight as taught by Sahin.
One would be motivated to make the modification such that the system can ensure gradual power increases as taught by Sahin; Paragraph 66.
Regarding claims 10 and 39, Jeon teaches the resources span a configured range in time/frequency (the resources are with respect to time/frequency resources; Paragraph 202. Further, resources allocated OTA are with respect to time and/or frequency which would thus be a configured range. As the “Configured range” is not defined, as long as resources are being allocated with respect to time and/or frequency, the prior art properly reads on the claim language).
Regarding claims 11 and 40, Jeon teaches transmitting an additional indication of UE capabilities (the base station receives, from the UEs, a local parameter updated based on the weight compression method; Paragraph 191. This is viewed as an additional parameter/indication sent.
Jeon does not expressly disclose non-coherent OTA. Sahin teaches federated learning for over-the-air aggregated computations using non-coherent detectors (Paragraph; 2, 82 and 84).
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include non-coherent OTA systems as taught by Sahin.
One would be motivated to make the modification such that the system can ensure gradual power increases as taught by Sahin; Paragraph 66.
Regarding claims 13 and 42, Jeon does not disclose the transmit power is associated with a pathloss of the channel for the UE; however, Sahin teaches that a weighting function can be used which is for the power of the transmitted OFDM symbols. A gradual power increase is tied directly to the magnitude of the local gradient; Paragraph 66. Further, Sahin teaches that path loss and power are taken into account; Paragraph 125. Thus as there is path loss in the system, the path loss would directly impact the power values. Everything that occurs in a system is “associated” with everything else, thus when combined with Jeon, the power would be associated with the sequence of Jeon.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include non-coherent OTA systems with transmission power based on the magnitude of the gradient as taught by Sahin.
One would be motivated to make the modification such that the system can ensure gradual power increases as taught by Sahin; Paragraph 66.
Regarding claim 14, Jeon teaches antennas to transmit the indication of the gradient update (The devices include one or more transceiver and/or antennas; Paragraph 198. Thus information being sent/received from these devices (including the indication of claim 1) would be sent via the antennas).
Claim(s) 2, 3, 31, 32 is/are rejected under 35 U.S.C. 103 as being unpatentable over Jeon in view of Sahin and further in view of Raghothaman et al. “Raghothaman” US 2019/0239083.
Regarding claims 2 and 31, the prior art does not disclose RRC, MAC-CE, DCI or SI carrying configuration information; however, Raghothaman teaches the use of DCI for sending configuration information (Paragraph 103) and also using RRC configurations with respect to OTA communications; See Figures 3-7.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include using DCI or RRC for sending configuration information as taught by Raghothaman.
One would be motivated to make the modification such that the UE can be provided various OTA configurations through RRC for scheduling among other things as taught by Raghothaman; Paragraphs 90, 97 and see also Figures 3-7.
Regarding claims 3 and 32, the prior art teaches the use of non-coherent OTA with respect to sequences as shown above, but the prior art does not disclose a quantization level associated with transmit power. Raghothaman teaches updating a signature vector based on power and the vector is then quantized; Paragraph 49.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include quantizing the power levels as taught by Raghothaman.
One would be motivated to make the modification such that the system can create quantized vectors based on power for each UE such that a primary RP (radio point) can be established/determined for each UE as taught by Raghothaman; Paragraph 49.
Claim(s) 7, 36 is/are rejected under 35 U.S.C. 103 as being unpatentable over Jeon in view of Sahin and further in view of Jeon et al. “Jeon-1” US 2024/0322941.
Regarding claims 7 and 36, the prior art does not teach the power is associated with a local batch size of the UE; however, Jeon-1 teaches transmission power is associated with the batch size; Paragraph 166.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include the transmission power being associated with the batch size of the UE as taught by Jeon-1.
One would be motivated to make the modification such that the system can maximize the efficiency of the federated learning as taught by Jeon-1; Paragraph 166.
Claim(s) 9, 38 is/are rejected under 35 U.S.C. 103 as being unpatentable over Jeon in view of Sahin and further in view of Zhang et al. “Zhang” US 11,943,114.
Regarding claims 9 and 38, the prior art does not teach learning weights associated with the UE are based on assignments from the network node; however, Zhang teaches the use of federated training/learning and a base station assigns weights to information associated with the users; Column 8 Lines 52-55 and lines 61-64.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include the weights are assigned by a network node as taught by Zhang.
One would be motivated to make the modification such that the learning modules can make predictions with attention to a weighted federated learning procedure as taught by Zhang; Column 8 Lines 44-46.
Claim(s) 12, 41 is/are rejected under 35 U.S.C. 103 as being unpatentable over Jeon in view of Sahin and further in view of Hasegawa et al. “Hasegawa” US 2024/0295625.
Regarding claims 12 and 41, while Jeon teaches transmitting sequences with a length, the prior art does not expressly disclose these sequences are pseudo-random.
Hasegawa teaches the length and type of sequences being transmit are pseudo-random; Paragraph 243. The system is for federated learning; Paragraph 291.
Thus it would have been obvious to one of ordinary skill in the art at the time of the effective filing to modify the teachings of the prior art to include pseudo-random sequences as taught by Hasegawa.
One would be motivated to make the modification such that this information can be used during the training procedure as taught by Hasegawa; Paragraph 243.
Response to Arguments
Applicant's arguments filed 12/15/2025 have been fully considered but they are not persuasive.
Regarding claim 1, Applicant argues Sahin, in paragraphs 65-66, teaches a weighting function to choose the symbols and not a transmit power as in claim 1. Therefore, Sahin does not disclose a transmit power based on a magnitude of at least one gradient update.
The Examiner respectfully disagrees. The claim language only requires the transmit power associated with a sequence be partly based on a magnitude of the gradient update. The Examiner notes the claim language does not specify exactly how the correlations impact one another, but that you have a transmit power of a sequence and magnitude of a gradient update. Generally speaking, anything that happens in a system is “Associated with” and “Based on” everything else in the system. Therefore, without defining exactly what correlations are, the prior art need only to include these particular elements. Turning to Sahin, Sahin teaches, in paragraph 66, that when the magnitude of the gradient is large, there is a gradual power increase (transmission power). The weighting function can range from 0-1 to limit the power of the transmitted symbols as well. Thus one can see the transmission power is associated with/based on magnitude of the gradient as claimed.
Regarding claim 5, Applicant argues Sahin does not disclose the newly amended limitations. As shown in the updated rejection, Sahin teaches the value of p (in the equation of paragraph 66) is a non-negative number thus this would “convert” any negative gradient range to a non-negative range once the equation is completed.
Regarding claim 7, Applicant argues paragraph 166 of Jeon-1 does not teach the transmit power used by a UE is based on a local batch size associated with a UE but does not provide any specifics why the cited portion doesn’t read on the claim language. The Examiner respectfully disagrees. Jeon-1 teaches transmission power is associated with the batch size; Paragraph 166.
Regarding claim 8, the Applicant argues the prior art does not teach or suggest the transmit power is associated with a local learning weight based on a local training loss for the federated learning procedure and refers to the same rationale as with respect to claim 1.
The Examiner respectfully disagrees. Sahin teaches the power values are based on a weighted function (i.e. local learning weight); Paragraph 66. With respect to the federal learning model, Sahin teaches local loss function models; Paragraph 56 which is associated with the weighting function of paragraph 66. The Examiner notes, everything that occurs in a system can broadly be assumed to be “based on” everything else that happens in the system. Without defining the exact correlation (or how one impacts the other) there is no real meaning behind the “based on” terminology.
Regarding claim 13, Applicant argues the prior art does not teach or suggest the transmit power associated with the at least one sequence is further based on a pathloss associated with a channel for the UE because paragraph 125 teaches pathloss with respect to convergence and not transmit power. The Examiner respectfully disagrees.
Sahin teaches that a weighting function can be used which is for the power of the transmitted OFDM symbols. A gradual power increase is tied directly to the magnitude of the local gradient; Paragraph 66. Further, Sahin teaches that path loss and power are taken into account; Paragraph 125. Thus as there is path loss in the system, the path loss would directly impact the power values. Everything that occurs in a system is “associated” with everything else, thus when combined with Jeon, the power would be associated with the sequence of Jeon. Without express definition of how the path loss and power are tied together, the prior art properly rejects the broad claim limitations.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRANDON M RENNER whose telephone number is (571)270-3621. The examiner can normally be reached Monday-Friday 7am-5pm 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, Derrick Ferris can be reached at (571)-272-3123. 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.
/BRANDON M RENNER/Primary Examiner, Art Unit 2411
Prosecution Insights