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 .
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 09/17/2024 has been placed in record and considered by the examiner.
Summary
This action is in reply to Applicant’s Amendments and Remarks filed on 08/14/2024.
Claims 1-4, 6-8, 12, 14, 16-22, 34 and 36 are pending.
Claims 5, 9-11, 13, 15, 22-33 and 35 are canceled.
NOTICE for all US Patent Applications filed on or after March 16, 2013
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.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of AIA 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1-4, 6, 34 and 36 are rejected under 35 U.S.C. 102 (a)(1) as anticipated by Kundu et al. (“Channel Estimation for Reconfigurable Intelligent Surface Aided MISO Communications: From LMMSE to Deep Learning Solutions”, hereinafter ‘KUNDU’).
Regarding claim 1, KUNDU teaches an apparatus (
Page 472 Left Column Paragraph 3:
there is little work on the design of MMSE channel estimators for RIS-aided communications, with the exception of the recent contribution [24], which considered a configuration with a line-of-sight channel between the BS and RIS, known perfectly to the BS.
Page 472 Right Column, II System Model
a multiple-input single-output (MISO) communication system where a BS with M antennas serves a single antenna UE with the help of an RIS having K passive elements; see Fig. 1. The RIS elements can intelligently control the phase of the incoming electromagnetic wave.
(Construed an apparatus as controller for RIS elements is implicit, either separate or combined with BS)
Page 473, Figure 1, BS with RIS) comprising:
at least one processor, and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, with the at least one processor (
See above Page 472 and Figure 1
(Construed that the apparatus as controller for RIS elements, either separate or combined with BS, comprises at least a processor, a memory storing computer program code to be executed by the processor), cause the apparatus at least to:
train a machine learning model to learn a configuration matrix that defines a reconfigurable intelligent surface (
Page 471, I. Introduction Right Column – Page 472 Left Column:
A distinctive feature of RIS is that they do not have active components like power amplifiers and analog-to-digital converters, but instead are composed of almost passive elements that can intelligently directionally control the propagation of impinging electromagnetic waves to improve end-to-end performance.
Page 472 Left Column Paragraph 3:
configuration with a line-of-sight channel between the BS and RIS, known perfectly to the BS.
Page 472 Right Column, II System Model
a multiple-input single-output (MISO) communication system where a BS with M antennas serves a single antenna UE with the help of an RIS having K passive elements; see Fig. 1. The RIS elements can intelligently control the phase of the incoming electromagnetic wave.
Page 473 Figure 1, Paragraph 1:
During training, the phase shifts of the RIS elements can be configured to assist the UL channel estimation.
Page 485, Left Column Paragraph 2:
we have proposed two CNN-based image denoising machine learning architectures to approximate this solution ….
See also Page 479:
B. FFDNET ARCHITECTURE
The authors in [26] have proposed a CNN-based image denoising architecture, FFDNet, which provides better denoising performance by utilizing the noise variance information. We now present the architecture of FFDNet that utilizes the noise variance information to further improve the channel estimation performance.
Page 480. Figure 5, Right Column:
PNG
media_image1.png
200
400
media_image1.png
Greyscale
(The K passive elements configuration matrix is known to BS and RIS controller sperate or with BS, and FFDNET is trained offline for a machine learning model learning RIS surface configuration to be used at the BS for run time channel estimation));
configure the reconfigurable intelligent surface for channel estimation during runtime, using the learned configuration matrix (
Page 472 Left Column Paragraphs 3-4 – Right Column Paragraphs 1-2:
In this work, we consider the design of optimal channel estimation strategies for RIS-aided multiple antenna communications, based on the minimum mean squared error (MMSE) criterion…..
Here we consider a general scenario in which all channels (BS to RIS, RIS to UE, and BS to UE) are subject to Rayleigh fading and are unknown…… first we derive the best linear estimator that minimizes the mean squared error, i.e., the LMMSE ……. The LMMSE performance depends on the phase shifts employed at the RIS during channel estimation, which may be optimized. Finding the optimum phase shifts ….. we present an algorithm based on the majorization-minimization (MM) …. …include the analytical DFT phase-shift matrix that has been shown to be optimal for the case of LS channel estimators.
We then move beyond the class of linear filters to present algorithms for approximating the optimal (non-linear) MMSE channel estimator for RIS-aided multiple-antenna communications. To this end, we introduce data-driven deep learning approaches. Specifically, we propose convolutional neural network (CNN)-based channel estimators that approximate the optimal MMSE solution. Our approach is to consider a linear LS channel estimate as a noisy 'image' at the neural network input, and then to apply a CNN based image denoising network to 'clean' this image and yield an improved RIS channel estimate.
Page 473 Paragraph 1:
….. the UL channel is estimated at the BS. During training, the phase shifts of the RIS elements can be configured to assist the UL channel estimation. The received pilot signal at the BS during the t-th training step is given by
PNG
media_image2.png
200
400
media_image2.png
Greyscale
Page 481 Right Column Paragraph 3:
Hence, it is important to study the robustness of the proposed CNN-based channel
estimators for scenarios where the statistical features of the channels used in ( online) testing differ from those used in ( offline) training.
(Construed RIS is configured during runtime using learned configuration matrix));
perform channel estimation on an uplink channel using the reconfigurable intelligent surface (
Page 472 Right Column Paragraph 2:
Our approach is to consider a linear LS channel estimate as a noisy 'image' at the neural network input, and then to apply a CNN based image denoising network to 'clean' this image and yield an improved RIS channel estimate.
See Page 473 Paragraph 1 and Eq. (1),
RIS elements can be configured to assist the UL channel estimation. The received pilot signal at the BS during the t-th training step and k phase shifts during i-th training based on
PNG
media_image3.png
200
400
media_image3.png
Greyscale
Page 473 Paragraph 2:
PNG
media_image4.png
200
400
media_image4.png
Greyscale
See also Page 478 Right Column
IV DEEP LEARNING APPROACHES TO MMSE CHANNEL ESTIMATION
In general, our approach is to introduce CNN-based estimators that take as input the sub-optimal LS estimate and produce as output an improved channel estimate that removes
noise from the LS estimate, and approximates the optimal MMSE solution.
Page 479 Right Column:
B. FFDNET ARCHITECTURE
The authors in [26] have proposed a CNN-based image denoising architecture, FFDNet, which provides better denoising performance ….
the input to the FFDNet is the noisy LS channel estimate of size M x (K + 1) x 2, ….
Page 480 Left Column:
PNG
media_image5.png
200
400
media_image5.png
Greyscale
); and
reconfigure the reconfigurable intelligent surface after the channel estimation to improve coverage within the uplink channel (
Page 471, I. Introduction Right Column:
RIS can potentially improve network coverage, and can enable spectral and energy efficiency gains that are commensurate with massive MIMO systems.
Page 484 Right Column Paragraphs 2-3:
The performance improvement of the DnCNN and FFDNet based estimators over the linear estimators is most evident for smaller numbers of RIS elements, K.
It is observed that first the achievable rate increases with K due to the increasing beamforming gain from the RIS elements (proportional to K2 ); however, after reaching a maximum value, the achievable rate starts decreasing. This is because, at large values of K, the channel estimation overhead penalty in the pre-log factor of (57) dominates the beamforming gain.
Page 485 Right Column Paragraph 2:
The channel estimates produced by the estimators that we propose can be employed in the implementation of passive and active beamforming vectors…..).
Regarding claim 2, KUNDU teaches the apparatus of claim 1, wherein the uplink channel comprises a cascaded uplink channel from a user equipment to the reconfigurable intelligent surface, and then from the reconfigurable intelligent surface to a network node (
Page 473 Figure 1, Paragraph 2:
PNG
media_image4.png
200
400
media_image4.png
Greyscale
See also Page 484 Table 2:
Cascaded Channel MSE).
Regarding claim 3, KUNDU teaches the apparatus of claim 1, wherein training the machine learning model comprises meeting a target normalized mean square error for at least one channel coefficient estimate (
Page 472 Left Column Paragraphs 3-4 – Right Column Paragraphs 1-2:
In this work, we consider the design of optimal channel estimation strategies for RIS-aided multiple antenna communications, based on the minimum mean squared error (MMSE) criterion…..
Here we consider a general scenario in which all channels (BS to RIS, RIS to UE, and BS to UE) are subject to Rayleigh fading and are unknown…… first we derive the best linear estimator that minimizes the mean squared error, i.e., the LMMSE ……. The LMMSE performance depends on the phase shifts employed at the RIS during channel estimation, which may be optimized. Finding the optimum phase shifts ….. we present an algorithm based on the majorization-minimization (MM) …. …include the analytical DFT phase-shift matrix that has been shown to be optimal for the case of LS channel estimators.
Page 478 Right Column
IV DEEP LEARNING APPROACHES TO MMSE CHANNEL ESTIMATION
In general, our approach is to introduce CNN-based estimators that take as input the sub-optimal LS estimate and produce as output an improved channel estimate that removes
noise from the LS estimate, and approximates the optimal MMSE solution.).
Regarding claim 4, KUNDU teaches the apparatus of claim 1, wherein the reconfigurable intelligent surface comprises a plurality of passive elements without a radio frequency part (
Page 472 Right Column, II System Model
a multiple-input single-output (MISO) communication system where a BS with M antennas serves a single antenna UE with the help of an RIS having K passive elements; see Fig. 1.).
Regarding claim 6, KUNDU teaches the apparatus of claim 1, wherein training the machine learning model comprises:
determining a number of time instants, the number of time instants being less than a number of elements within the reconfigurable intelligent surface (
Page 472 Right Column, II System Model
a multiple-input single-output (MISO) communication system where a BS with M antennas serves a single antenna UE with the help of an RIS having K passive elements; see Fig. 1.
Page 473 Paragraph 1:
….. the UL channel is estimated at the BS. During training, the phase shifts of the RIS elements can be configured to assist the UL channel estimation. The received pilot signal at the BS during the t-th training step is given by
PNG
media_image6.png
200
400
media_image6.png
Greyscale
Page 474 Left Column Paragraph 5:
i = 1, 2, .. . , K
(Construed Number of Training Time Instant I less than or equal RIS elements K)).
Regarding claim 34, the claim is interpreted mutatis mutandis of claim 1 and rejected for the same reason as set forth for claim 1.
Regarding claim 36, the claim is interpreted mutatis mutandis of claim 1 and rejected for the same reason as set forth for claim 1.
Allowable Subject Matter
Claims 7-8, 12, 14, 16-22 are 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 in intervening claims.
The following is a statement of reasons for the indication of allowable subject matter:
Regarding claim 7, KUNDU or any other prior art of record either alone or in combination fails to teach the apparatus of claim 6, wherein training the machine learning model further comprises:
initializing the configuration matrix to be a truncated discrete Fourier transform matrix;
obtaining a dataset of a plurality of cascaded channel matrices;
selecting, in an epoch, a random mini-batch of channel matrices; and
determining a received uplink signal corresponding to one of the cascaded channel matrices within the random mini-batch of channel matrices.
Regarding claims 8, 12, 14, 16-21, the claims being dependent on claim 7, are also interpreted same as claim 7.
Regarding claim 22, KUNDU or any other prior art of record either alone or in combination fails to teach the apparatus of claim 1, wherein training the machine learning model comprises:
determining a number of time instants, the number of time instants being less than a number of elements within the reconfigurable intelligent surface;
initializing the configuration matrix to be a truncated discrete Fourier transform matrix;
obtaining a dataset of a plurality of cascaded channel matrices;
selecting, in an epoch, a random mini-batch of channel matrices;
determining a received uplink signal corresponding to one of the cascaded channel matrices within the random mini-batch of channel matrices;
determining a prediction of the one of the cascaded channel matrices;
computing a normalized mean square error for the mini-batch;
determining whether a stopping criterion is reached;
in response to the stopping criterion being reached, quantizing the entries of the configuration matrix to a nearest permissible value; and
in response to the stopping criterion not being reached, performing a gradient descent on the configuration matrix and beginning a new epoch.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Zhou et al. (US 12646835 B2), describing Electronic Device, Wireless Communication Method And Computer-readable Storage Medium
Elshafie et al. (US 20240365152 A1), describing RIS CONFIGURATION COMPUTATION USING REINFORCEMENT LEARNING
Medra et al. (US 20250047328 A1), describing SYSTEMS AND METHODS FOR ROBUST BEAMFORMING USING A RECONFIGURABLE INTELLIGENT SURFACE IN COMMUNICATION SYSTEMS
Taniguchi et al. (US 20260113075 A1), describing RADIO WAVE PROPAGATION ENVIRONMENT REPRODUCTION SYSTEM AND RADIO WAVE PROPAGATION ENVIRONMENT REPRODUCTION METHOD
Sun et al. (US 20230129288 A1), describing METHOD AND DEVICE FOR RIS AUTOMATIC SETUP
Mao et al., “Channel Estimation for Intelligent Reflecting Surface Assisted Massive MIMO Systems-A Deep Learning Approach"
Zhang et al, “Deep Learning Optimized Sparse Antenna Activation for Reconfigurable Intelligent Surface Assisted Communication”
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHAH M RAHMAN whose telephone number is (571)272-8951. The examiner can normally be reached 9:30AM-5:30PM PST.
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, UN C CHO can be reached at 571-272-7919. 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.
/SHAH M RAHMAN/Primary Examiner, Art Unit 2413
Prosecution Insights