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
Applicants’ arguments filed on 4 August 2026 have been fully considered but they are not deemed to be persuasive.
By the amendment filed 4 August 2026, claims 1, 4, 7, 8, 12, 13, and 20 have been amended; claims 6 and 17 have been cancelled.
Claims 1-5, 7-16, and 18-20 are now pending.
Claims _ are rejected.
Response to Arguments
Applicant’s arguments concerning amended claim 1 have been considered but are not persuasive.
Regarding “sending information about a third model to a terminal device, wherein input of the third model is information about a sounding signal, and output of the third model is characteristic information of a channel,” applicant argues that Zhu’s conventional channel estimator differs from a trained neural network whose structure or parameters are transmitted to and deployed at the terminal. However, claim 1 does not expressly require transmitting a complete neural network, transmitting trained weights, training the third model at the network, or supplying raw sounding-signal samples directly to a neural network. Rather, the claim recites the broader expressions “information about a third model,” “information about a sounding signal,” and “characteristic information of a channel.” These expressions receive their broadest reasonable interpretation consistent with the specification. Restrictions concerning a particular training or deployment architecture should not be imported into the claim merely because applicant’s arguments describe that architecture. See MPEP § 2111.
Kaya discloses transmitting measurement-report configuration information from the base station to the terminal, including information identifying reference-signal resources to be measured, the number of resources or beams to be reported, and reporting timing or frequency. The terminal processes the selected reference signals and provides resulting channel-related measurements to the base station. Kaya further identifies channel-information quantities including CQI, PMI, CRI, RI, and L1-RSRP. See Kaya, ¶¶[0039]–[0041]. Kaya also discloses sending updated report configurations based on the predicted future beam sequence. See Kaya, ¶[0052]. Zhu discloses terminal-side channel estimator 358 deriving channel estimates from a reference signal or feedback transmitted by the base station. See Zhu, ¶¶[0045], [0048]. Thus, Zhu teaches processing reference-signal information to produce information characterizing a channel, while Kaya teaches network-provided configuration governing terminal-side reference-signal measurement and reporting.
The rejection relies on applying Kaya’s network-provided configuration to Zhu’s terminal-side channel-estimation processing. It would have been obvious to configure that processing according to the reference-signal resources and measurement requirements selected by the base station so that the resulting channel information corresponds to the resources required for network-side beam selection. In this combination, the transmitted configuration specifies the reference-signal inputs and operating conditions governing the channel-estimation model. Such configuration constitutes information about the operation of that model; the claim does not require that the transmitted information reproduce the model’s complete structure or parameters. Likewise, reference-signal measurements constitute information about the sounding signal, and the resulting channel estimates constitute characteristic information of the channel.
The reason for the modification follows from the complementary functions disclosed by the references. Kaya’s base station selects reference-signal resources and reporting requirements, and Zhu’s terminal-side estimator derives channel information from received reference signals. Applying the selected configuration to the estimator would coordinate terminal-side channel characterization with the network’s use of the resulting measurements. The modification retains the disclosed functions of network-directed measurement configuration and terminal-side channel estimation and provides the predictable result of channel information corresponding to the configured reference-signal resources.
Furthermore, both Kaya and Zhu expressly disclose neural-network processing. Kaya describes a neural-network beam-sequence model and training that model by adjusting its weights based on prediction error. See Kaya, ¶¶[0053]–[0054]. Zhu describes a terminal-side artificial neural network having neural input nodes, hidden nodes, and output nodes that process beam-related information and predict beam conditions. Its inputs include historical beam-pair strengths, beam indices, and angles of departure and arrival, and its outputs include predicted sets of good and bad beam pairs. See Zhu, ¶¶[0065]–[0075]. Zhu additionally discloses the UE training and executing the artificial neural network. See Zhu, ¶¶[0081]–[0082]. Accordingly, applicant’s characterization of Zhu’s Fig. 3 estimator as conventional does not address Zhu’s broader disclosure or establish that the combined teachings exclude model-based processing. The Office does not equate Zhu’s Fig. 3 estimator with its Fig. 7 neural network; rather, these disclosures demonstrate terminal-side channel estimation and terminal-side neural-network processing, respectively.
Applicant’s observation that Kaya’s beam-sequence model operates at the network likewise does not defeat the combination. The proposed modification does not require physically relocating Kaya’s beam-sequence model to the terminal. It uses Kaya’s network-directed configuration with Zhu’s terminal-side processing. Accordingly, the absence of an express disclosure of downloading a complete trained neural network does not distinguish the broadly recited limitation, and applicant’s arguments do not overcome the rejection of amended claim 1 over Kaya in view of Zhu.
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.
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 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.
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.
Claims 1-5, 7-16 and 18-22 are rejected under 35 U.S.C. § 103 as being unpatentable over Kaya et al. (US 2022/0190883 A1) in view of Zhu et al. (US 2019/0191425 A1).
Regarding claim 1, Kaya discloses a communication method, comprising:
receiving characteristic information of a downlink channel from the terminal device; Kaya discloses receiving terminal measurement reports containing reference-signal measurements, including RSRP and associated beam or resource identifiers. Kaya identifies channel-state information as including CQI, PMI, CRI, SSBRI, LI, RI, and L1-RSRP. These reports provide information characterizing the downlink channel. See Kaya, ¶¶[0038]–[0041], [0072].
obtaining first codeword information based on the characteristic information of the downlink channel and a first model, wherein input of the first model comprises the characteristic information of the downlink channel, output of the first model comprises the first codeword information, and a first codeword corresponding to the first codeword information is used for precoding. Kaya discloses deriving a past beam sequence from terminal measurement reports and supplying that sequence to a beam-sequence model to predict future beam indices. The reported beam indices identify beams associated with measured channel conditions, and the predicted indices identify beams for subsequent communication. See Kaya, ¶¶[0051], [0053]–[0054], [0072]–[0075]. Kaya further explains that a beam is formed by applying a set of beam weights, including gain and/or phase, to the antenna system. Thus, the predicted beam index provides information identifying the corresponding beamforming weight set used for precoding. See Kaya, ¶[0032].
Kaya further discloses transmitting configuration information to the terminal specifying reference-signal resources to be measured, the number of resources or beams to be reported, and reporting timing or frequency. The terminal uses this configuration to obtain and report channel-related measurements. See Kaya, ¶¶[0039]–[0041]. Kaya also discloses updating that configuration based on the model’s predicted beam sequence. See Kaya, ¶[0052].
Kaya does not expressly identify the configured terminal-side reference-signal processing as a third model having the claimed input/output relationship.
Regarding sending information about a third model to a terminal device, wherein input of the third model is information about a sounding signal, and output of the third model is characteristic information of a channel; Zhu discloses terminal-side channel estimator 358, which derives channel estimates from a reference signal or feedback transmitted by the base station. See Zhu, ¶¶[0045], [0048]. The reference signal probes the channel and supplies information from which the estimator determines channel characteristics. Under the broadest reasonable interpretation consistent with the specification, the estimator provides a processing model whose input includes information about a sounding signal and whose output includes characteristic information of the channel. The claim does not expressly require raw sounding-signal samples, a particular channel-information representation, or a trained neural-network implementation.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply Kaya’s network-provided reference-signal measurement configuration to Zhu’s terminal-side channel-estimation processing. Specifically, the base station would transmit configuration identifying the reference-signal resources and measurement conditions governing the estimator’s operation. In the resulting combination, that configuration would constitute information about the operation of the third model, including the reference-signal inputs on which its channel-characterization processing is performed.
A person of ordinary skill would have been motivated to make this modification to coordinate the terminal’s channel-estimation processing with the reference-signal resources and reporting requirements selected by the base station, thereby providing channel information corresponding to the resources used for network-side beam selection. The modification combines Kaya’s network-directed measurement configuration with Zhu’s terminal-side reference-signal channel estimation, retaining their respective functions and producing the predictable result of channel characterization for the configured resources.
Accordingly, Kaya in view of Zhu teaches or suggests the communication method of amended claim 1, including transmission of configuration information governing a terminal-side model that processes sounding-signal information to obtain channel-characteristic information.
Regarding claim 2, Kaya further discloses wherein the method further comprises:
precoding first downlink data based on the first codeword, to obtain precoded data; andKaya discloses beamforming based on the selected beam/codeword, which inherently includes precoding downlink data. (¶[0069, 72])
sending the precoded data to the terminal device.Kaya discloses transmitting signals to the terminal using the selected beam (¶[0069]).
Regarding claim 3, Kaya does not specifically disclose sending K sounding signals to the terminal device based on K sounding signal codewords, wherein each of the K sounding signal codewords is used to send one corresponding sounding signal of the K sounding signals.
Zhu discloses sending sounding/reference signals and associating them with beams/codewords, as evidenced by:
“measure each beam pair from multiple available beam pairs…” (Zhu, ¶[0008])and the disclosure of “Reference Signal” and “Channel Estimator” (Zhu, Fig. 3).
Thus, it would have been obvious to send K sounding signals… each corresponding to a respective codeword, since Zhu teaches transmitting reference signals across beams to obtain channel measurements, which would improve the model-based beam selection of Kaya.
Regarding claim 4, Kaya discloses … determining codeword information based on characteristic information of a respective channel, and a codeword corresponding to the codeword information is used for precoding.
Kaya discloses determining codeword information based on characteristic information of a channel using a model, as evidenced by:
“beam sequence model… measurement reports… estimating the quality… of different beams” (¶[0072]).
Kaya-Zhu does not explicitly disclose wherein the first model is comprised in M₁ first-type models, M₁ is an integer greater than 1, each of the M₁ first-type models corresponds to a respective channel type.
However, Kaya discloses training based on diverse channel conditions using measurement reports (¶[0072]).
Thus, it would have been obvious to implement M₁ first-type models corresponding to respective channel types, since model specialization based on channel conditions improves prediction accuracy and is well known in the art.
Regarding claim 5, Kaya discloses wherein the first model is obtained through training based on a first training dataset, the first training dataset comprises one or more downlink channel samples, and the first training dataset corresponds to a first channel type, as evidenced by:
“training phase… substantial amount of training data” (¶[0072]).
Kaya further discloses the first training dataset comprises one or more downlink channel samples, as the training data includes measurement reports representing channel conditions (¶[0072]).
Claim 7 recites wherein the third model is comprised in M1 second-type models, M1 is an integer greater than 1, the Mi first=type models are in a one=to-one correspondence with the Mi second=type models, input of each of the M1 second-type models is information about a sounding signal, and output of each of the M1 second-type models is characteristic information of a channel, wherein M1 is an integer greater than 1. Kaya discloses model-based processing as recited above.
Kaya discloses all of the subject matter except multiple models for sounding-based estimation.
Zhu discloses channel estimation under different conditions using reference signals. (Fig. 3)
Thus, it would have been obvious to use multiple models for channel estimation to improve robustness across varying channel conditions.
Regarding claim 8, Kaya discloses wherein the third model is obtained through training based on a first training dataset…, the first training dataset comprises one or more downlink channel samples, and the first training dataset corresponds to a first channel type, as evidenced by:
“training phase… substantial amount of training data for the beam sequence model” (Kaya, ¶[0072]),
“measurement reports… for estimating the quality… of different beams” (Kaya, ¶[0072]).
Kaya does not explicitly disclose training the model based on assistance information of the terminal device.
Zhu discloses using terminal-related information and measurements (e.g., beam measurements and reference signal processing) to determine channel characteristics, as evidenced by:
“measure each beam pair… based on the measurement results” (Zhu, ¶[0008])and the use of Reference Signal and Channel Estimator (Zhu, Fig. 3).
Thus, it would have been obvious to train the model based on both channel samples and assistance information of the terminal device, since incorporating additional terminal-related information (e.g., measurements, context) improves model accuracy and channel estimation performance.
Regarding claim 9, Kaya discloses collecting measurement reports across different UE positions and movement conditions, as evidenced by:
UE traveling along roads with different positions and conditions (Kaya, ¶[0069]).
Kaya does not explicitly disclose assistance information including location, speed, acceleration, posture, or movement direction information.
Zhu discloses using measurement-based information associated with UE movement and beam conditions (¶[0008]).
Thus, it would have been obvious to include location, speed, acceleration, posture, or movement direction information as assistance information, since such information directly impacts channel characteristics and beam selection and would improve model performance.
Regarding claim 10, Kaya discloses channel conditions affected by UE position, obstacles, and propagation environment, as evidenced by:
“an obstacle… may cause a blocking condition… best beam… may be provided by [another TRP]” (Kaya, ¶[0069]).
Kaya does not explicitly disclose using assistance information to determine path information and then determining direction or coverage width.
Zhu discloses determining beam pairs and selecting beams based on measurement results corresponding to channel conditions (¶[0008]).
Thus, it would have been obvious to use assistance information to determine path characteristics (e.g., direction, coverage width) and then determine beam/codeword accordingly, since channel path characteristics directly influence beam direction and coverage.
Regarding claim 11, Kaya discloses interaction with the terminal device including collection of measurement reports and system configuration for model-based processing (¶[0072]).
Kaya does not explicitly disclose receiving capability information including neural network support, computing power, storage capability, or sensor configuration.
Zhu discloses system components and processing structures including neural network processing and signal handling (Fig. 3; Fig. 9).
Thus, it would have been obvious to receive capability information of the terminal device, since adapting processing (e.g., model selection, computation distribution) based on terminal capabilities is a well-known design consideration to improve efficiency and compatibility.
Regarding claim 12, the claimed method recites terminal-side operations corresponding to the network-side configuration and channel-information exchange addressed in claim 1, together with receiving K sounding signals and obtaining channel-characteristic information therefrom. These operations are taught or suggested by Kaya, ¶¶[0038]–[0041], [0052], and Zhu, ¶¶[0045], [0048], [0053]–[0054]. Accordingly, claim 12 is analogous to claim 1 and is similarly rejected under 35 U.S.C. § 103 over Kaya in view of Zhu for the corresponding mappings and combination rationale stated above.
Regarding claim 13, Kaya discloses inputting channel-related measurement information into a model, as evidenced by “training data… measurement reports… beam sequence model” (Kaya, ¶[0072]). Zhu discloses obtaining channel information based on sounding signals (Zhu, ¶[0008]; Fig. 3). Thus, it would have been obvious to input sounding signal information, and optionally assistance information, into a model to obtain channel characteristics, since combining Zhu’s sounding-based estimation with Kaya’s model-based processing yields predictable results.
Regarding claim 14, Kaya discloses that channel conditions vary with UE position and movement, as evidenced by UE movement and changing beam conditions (Kaya, ¶[0069]). Zhu discloses beam measurement under varying conditions (Zhu, ¶[0008]). Thus, it would have been obvious to include location, speed, acceleration, posture, or movement direction information as assistance information, since such parameters directly affect channel characteristics and beam selection.
Claim 15 recites wherein after the sending the first characteristic information to the access network device, the method further comprises: cyclically performing the following steps when i is an integer from 1 to N, wherein N is an integer:receiving an ith sounding signal from the access network device; obtaining an ith piece of characteristic information based on information about the ith sounding signal and a fourth model, wherein the ith piece of characteristic information is characteristic information of the downlink channel; and sending the ith piece of characteristic information to the access network device. The combination of Zhu and Kaya discloses these limitations. Specifically, Zhu further discloses repeated measurement of multiple beam pairs based on received signals (Zhu, ¶[0008]). Kaya further discloses repeated model-based processing of incoming measurement data (Kaya, ¶[0072]). Thus, it would have been obvious to cyclically process sounding signals using a model and send corresponding channel characteristics, since iterative estimation improves accuracy.
Claim 16 recites wherein the obtaining an ith piece of characteristic information based on information about the ith sounding signal and a fourth model comprises: inputting the information about the ith sounding signal into the fourth model, to obtain the ith piece of characteristic information; or inputting the information about the ith sounding signal and [[the]] assistance information of [[the]] a terminal device into the fourth model, to obtain the ith piece of characteristic information. The combination of Kaya and Zhu discloses these limitations. Specifically, Kaya discloses inputting measurement data into a model (¶[0072]). Zhu discloses deriving channel information from sounding signals (¶[0008]; Fig. 3). Thus, it would have been obvious to input sounding signal information, and optionally assistance information, into a model.
Regarding claim 18, the combination of Zhu and Kaya discloses wherein the method further comprises: sending channel estimation information to the access network device, wherein the channel estimation information is used to train the third model (Zhu, Fig. 3 “Channel Estimator”, channel estimation based on reference signals; Kaya, (¶[0072], training a model using channel-related data, as evidenced by “training phase… substantial amount of training data”). Thus, it would have been obvious to send channel estimation information for model training.
Regarding claim 19, the combination of Kaya and Zhu discloses wherein the method further comprises: sending capability information of a terminal device to the access network device, wherein the capability information indicates one or more of the following: whether the terminal device supports a neural network model, a type of a neural network model supported by the terminal device, computing power information of the terminal device, or storage capability information of the terminal device (Kaya, ¶[0072], use of neural network models and system considerations for processing; Zhu, Fig. 3 discloses system architecture for signal processing). Thus, it would have been obvious to provide capability information to enable appropriate model selection and processing.
Claim 20 recites an apparatus, comprising: a processor; and a memory storing instructions that, when executed by the processor, cause the processor to perform the method according to claim 12, and is thus similarly rejected.
Regarding claim 21, the apparatus according to claim 20, wherein the obtaining first characteristic information based on information about the K sounding signals and the third model comprises: inputting the information about the K sounding signals into the third model, to obtain the first characteristic information; or inputting the information about the K sounding signals and assistance information of a terminal device into the third model, to obtain the first characteristic information. Kaya in view of Zhu teaches or suggests the apparatus of claim 20. Zhu further discloses terminal-side channel estimator 358 deriving channel estimates from reference signals transmitted by the base station. Zhu also discloses transmitting multiple beams in a beam-reference-signal cycle and measuring candidate transmit-receive beam pairs. See Zhu, ¶¶[0045], [0048], [0053]–[0054], and Figs. 3, 5A–5B. Thus, the reference-signal information is supplied to channel-estimation processing to obtain information characterizing the channel. Applying that processing to the K configured sounding signals would have been obvious to obtain channel information for the respective configured resources, consistent with Kaya’s configuration of multiple reference-signal resources for terminal measurement. See Kaya, ¶¶[0039]–[0041].
Claim 21 recites the two processing alternatives disjunctively. The foregoing combination teaches or suggests the first alternative, inputting the information about the K sounding signals into the third model, to obtain the first characteristic information, which does not require terminal assistance information.
Regarding claim 22, the apparatus according to claim 21, wherein the assistance information of the terminal device comprises one or more of the following: location information of the terminal device, speed information of the terminal device, acceleration information of the terminal device, posture information of the terminal device, or movement direction information of the terminal device. Zhu discloses supplying terminal self-rotation angles and angular speed to neural input node 712, and terminal movement directions and moving speed to neural input node 714. These inputs are obtained using terminal gyroscopes and an accelerometer. Zhu further supplies historical beam-pair strengths, beam indices, and angles of departure and arrival to neural input node 716. See Zhu, ¶¶[0066]–[0068]. The terminal movement direction and moving speed expressly correspond to at least two of the recited alternatives.
It would have been obvious to supplement the sounding-signal information supplied to the terminal-side model of Kaya in view of Zhu with Zhu’s terminal movement-direction or speed information. Zhu teaches using these inputs together with beam-related measurements to account for terminal movement when predicting beam conditions, including changes in line-of-sight conditions and suitable beam pairs. See Zhu, ¶¶[0071]–[0075]. A person of ordinary skill would have been motivated to use the same movement information in the combined model’s channel-characterization processing to account for changes in the terminal’s position relative to the measured beams and thereby provide channel information responsive to terminal movement. The resulting combination would supply both sounding-signal information and terminal assistance information to the third model, as recited in claims 21–22.
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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to LUAT T PHUNG whose telephone number is (571)270-3126. The examiner can normally be reached on M-F 9 AM - 6 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, Marcus Smith can be reached on (571) 272-3988. 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). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Luat Phung/
Primary Examiner, Art Unit 2468
Prosecution Insights