CTNF 18/361,279 CTNF 99652 DETAILED ACTION Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia 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 Applicant’s submission, dated 10 March 2025—in which claims 1-11, 15-17, and 19 are amended, new claim 20 is added, and claims 1-19 are pending—has been entered into the record and is fully considered herein. Response to Arguments Applicant’s arguments—set forth at § I of Applicant's Remarks—with respect to claims 1 and 15, have been fully considered but the arguments are not found to be persuasive. In response to Applicant's argument that the references fail to show certain features of the invention, it is noted that the feature upon which Applicant relies (e.g., “using a cumulative distribution function (CDF)”), is not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns , 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). That is, the rejected claims do not positively recite any CDF processing by the terminal/base station. Therefore, Applicant’s arguments regarding CDF operations at the mobile terminal/base station are not commensurate with the scope of the claims. For example, Applicant asserts: Claim 1 has been amended to include the feature wherein, in embodiments corresponding to amended claim 1, the terminal quantizes a precoding vector using a cumulative distribution function (CDF). Embodiments according to claim 1 as amended are characterized in that the terminal quantizes element values of the precoding vector using the CDF. Here, the CDF may be a function defined based on element values of all dimensions of the transformed precoding vector. In addition, the CDF may also be a function defined respectively for the element values of each dimension of the transformed precoding vector. Applicant similarly asserts: Thus, the terminal may perform quantization using the CDF, and may apply quantization to values normalized to the range of 0 to 1, regardless of the distribution of the input values. Accordingly, even when the element values of the precoding vector follow any statistical form—such as a normal distribution, an asymmetric distribution, or a distribution with skew—the application of the CDF allows them to be transformed into values having a uniform probability distribution. By using such CDF-based quantization, the quantization intervals may be configured so that each quantization level is used with an equal probability, thereby providing the advantage of significantly reducing quantization noise. Applicant further asserts that “Therefore, none of the cited prior art, including NERINI, discloses or suggests the CDF- based element-value transformation and quantization procedure of the embodiments corresponding to present amended claim 1, which constitutes a clear and distinguishing technical feature over the prior art.” The Examiner notes, however, that claim 1 has been amended to delete the previously-recited limitation “through a first distribution function.” Manner of Making Amendments under 37 C.F.R 1.121 Applicant’s submission is objected to because the markings used to identify changes in the amended claims do not comply with MPEP § 714(II)(C). Specifically, several deletions are indicated by strike-through where double brackets are required. Any subsequent submission should include a listing of claims in which all claim changes are properly identified relative to the immediate prior version of the claims in accordance with MPEP § 714(II)(C). For example, claim 1 includes or more instances of “ s ” which cannot be easily perceived, and which instead should have been indicated as “[[s]]”. Similarly, claim 3 includes an instance of “ 2 ” which, for purposes of clarity, instead should have been indicated as “[[2]]”. These instances are exemplary and other claims may include additional improper deletion notations. Claim Rejections - 35 USC § 112 The following is a quotation of 35 U.S.C. § 112(b): (b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention. Claim 1 recites the limitation "receiving, from a base station” and the limitation “receiving a reference signal from a base station.” It is unclear whether claim 1 requires 2 distinct base stations. Accordingly, appropriate correction is required. Claim Objections The status identifier of one or more claims is not in accordance with 37 CFR 1.121 which requires the claim listing to identify the status of every claim after its claim number. In particular, claim 18 in the listing of claims does not include a status identifier. Accordingly, appropriate correction is required in any subsequent submission. Regarding claim 3, the claim term “parameter(s)” is improper because the notation “(s)” attempts to encompass both singular and plural alternatives within a single term. Accordingly, appropriate correction is required. Allowable Subject Matter 12-151-07 AIA 07-97 12-51-07 Claim s 11 and 12 are allowed. 12-151-08 AIA 07-43 12-51-08 Claim 20 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. Claim Rejections - 35 USC § 103 07-20-aia AIA The following is a quotation of 35 U.S.C. § 103 which forms the basis for all obviousness rejections set forth in the 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. 07-23-aia AIA 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. 07-21-aia AIA Claims 1-3, 8-10, 15, 17, and 18 are rejected un der 35 U.S.C. § 103 as being unpatentable over “Machine Le arning-based CSI Feedback with Variable Length in FDD Massive MIMO,” arXiv:2204.04723v1 , [cs.IT], Apr. 10, 2022, (hereinafter, “NERINI”) in view of US 2019/0173607 (hereinafter, “LIU”), and further in view of US 2024/0275462 (hereinafter, “GAO”). Regarding c laim 1, NERINI discloses: A method of a terminal ( UE ) , comprising: . . . generating at least one precoding vector . . . ; (§ III.B: After the offline learning stage, the first N P principal components, the vector μ train , and the codebook of the learned quantization levels are offloaded to the UE. This allows the online feedback process, as represented in Fig. 1. The UE firstly vectorizes the noisy downlink channel matrix ) generating transformed precoding vectors by performing transformation on the at least one precoding vector . . . ; ( Id. : Then, it compresses h DL = h ~ DL − μ train by considering only the first NP principal components, obtaining z DL ) quantizing the transformed precoding vectors; and ( Id. : [T]he compressed CSI z DL is quantized according to the learned codebook ) transmitting the quantized transformed precoding vectors to the base station. ( Id. : [T]he feedback transmission, [in which] the feedback is fully described by B bits, which are sent from the UE to the BS ) NERINI does not explicitly disclose: receiving, from a base station, information on spatial domain basis vectors and frequency domain basis vectors; using the spatial domain basis vectors and frequency domain basis vectors; receiving a reference signal from a base station; generating a precoding vector based on the reference signal; In the same field of endeavor, however, GAO teaches: receiving, from a base station, information on spatial domain basis vectors and frequency domain basis vectors; (¶ 0060: [N]etwork device 2 is a device that can communicate with the terminal device 1, and may be a base station ; ¶ 0077: The network device sends configuration information to the terminal ; ¶ 0078: [C]onfiguration information may indicate one or more spatial domain beam basis vector groups and Q thresholds ; ¶ 0027: Configuration information indicating one or more spatial domain beam basis vector groups and Q thresholds is sent to a terminal device. Amplitudes and phases of spatial-frequency combination coefficients in M spatial-frequency combination coefficient vectors are received from the terminal device. The Q thresholds correspond one-to-one to spatial domain beam basis vectors in the one or more spatial domain beam basis vector groups. The M spatial-frequency combination coefficient vectors are determined based on L spatial domain beam basis vectors, K frequency domain basis vectors corresponding to each of the L spatial domain beam basis vectors, and a target precoding vector ) using the spatial domain basis vectors and frequency domain basis vectors; (¶ 0003: A precoding vector is formed by performing linear combination on a plurality of orthogonal spatial domain beam vectors ; ¶ 0075: [A] two-dimensional spatial-frequency combination coefficient corresponding to both a spatial domain basis vector and a frequency domain basis vector ; ¶ 0093: 204. The terminal device determines M spatial-frequency combination coefficient vectors based on the L spatial domain beam basis vectors, the K frequency domain basis vectors corresponding to each of the L spatial domain beam basis vectors, and a target precoding vector ; ¶ 0095: [A]fter selecting the K frequency domain basis vectors from the frequency domain basis vector set for each of the L spatial domain beam basis vectors, the terminal device may determine the M spatial-frequency combination coefficient vectors based on the L spatial domain beam basis vectors, the K frequency domain basis vectors corresponding to each of the L spatial domain beam basis vectors, and the target precoding vector ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s machine learning-based CSI feedback to provide receiving information on spatial domain basis and frequency domain basis vectors as taught by GAO to provide transforming precoding vectors and quantizing the precoding vectors, such that indication information of the precoding vector is reported to the network device, so as to help the network device obtain an optimal precoding vector. See GAO, at ¶ 0069. Also, in the same field of endeavor, however, LIU teaches: receiving a reference signal from a base station; (¶ 0007: [R]eceiving, by a terminal, a downlink reference signal sent by an access network device ) generating a precoding vector based on the reference signal; (¶ 0008: [M]easuring, by the terminal, the downlink reference signal, to obtain a plurality of uplink precoding vectors ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s “offloading”—from the base station to the UE—of the information learned during its offline learning stage to provide for sending/receipt of a reference signal as taught by LIU to obtain precoding vectors such that—based on a channel reciprocity principle—the precoding vector is more suitable for the actual case of the uplink channel of the terminal, so that the transmission performance of the uplink data can be improved. See LIU, at ¶ 0120. Regarding claim 2, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 1. NERINI further discloses: further comprising: wherein in the generating of the transformed precoding vector, the transformed precoding vectors are generated by performing dimensionality reduction transformation on the at least one precoding vector through a principal component analysis scheme . . . . (§ I: [A] novel CSI feedback strategy based on PCA and k-means clustering. In this strategy, PCA is used to compress the channel matrix into a latent space with adaptive dimensionality, allowing to optimally design the feedback with variable length. To quantize this compressed channel, the feedback bits are smartly allocated to the principal components in order to minimize a properly defined distortion function. Finally, on each principal component, the quantization levels are determined with vector quantization ) NERINI does not explicitly disclose: using the spatial domain basis vectors and the frequency domain basis vectors. In the same field of endeavor, however, GAO teaches: using the spatial domain basis vectors and the frequency domain basis vectors. (¶ 0075: [A] two-dimensional spatial-frequency combination coefficient corresponding to both a spatial domain basis vector and a frequency domain basis vector ) Regarding claim 3, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 1. NERINI does not explicitly disclose: wherein information on the spatial domain basis vectors and the frequency domain basis vectors is received from the base station in a form of higher layer configuration parameter(s). In the same field of endeavor, however, GAO teaches: wherein information on the spatial domain basis vectors and the frequency domain basis vectors is received from the base station in a form of higher layer configuration parameter(s). (¶ 0077: The network device sends configuration information to the terminal ) Regarding claim 8, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 1. NERINI further discloses: wherein in the quantizing of the transformed precoding vector, elements of all dimensions of the transformed precoding vectors or elements of each dimension of the transformed precoding vector are quantized. (§ III.A: Since each column of Z train is a complex vector, each k-means clustering problem is considered as a vector quantization problem in the feature space R 2 . In this problem, the two dimensions are given by the real and imaginary parts of the Z train elements. Finally, the cluster centers of the n-th k-means clustering give the quantization levels for the n-th latent space dimension ) Regarding claim 9, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 8. NERINI further discloses: wherein when the elements of each dimension of the transformed precoding vectors are quantized, a bit length is maintained to be identical between dimensions or decreases as the dimension increases. (§ V.B: [T]he feedback bits are allocated uniformly to all the latent space dimensions, and the quantization levels are determined with k-means clustering ) Regarding claim 10, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 8. NERINI further discloses: wherein when the elements of each dimension of the transformed precoding vectors are quantized, a bit length for each dimension is determined according to importance based on an explained variance of each dimension. (§ I: [A]llocate the feedback bits to the latent space dimensions, i.e., the principal components, according to their importance ; § III.A: [T]he NP principal components have decreasing importance, defined as the variance explained by each of them. Thus, the optimal discretization approach is expected to allocate more bits to more important principal components, and fewer bits to the less important ones ) Regarding claim 15, NERINI discloses: A method of a base station ( base station (BS) ) , comprising: . . . receiving quantized transformed precoding vectors . . . ; (§ III.B: [T]he feedback is fully described by B bits, which are sent from the UE to the BS ) . . . reconstructing the at least one precoding vector from the transformed precoding vector. ( Id., [I]t reshapes hˆ DL + µ train into H DL , which is the downlink channel matrix estimate ) NERINI does not explicitly disclose: transmitting, to a terminal, information on spatial domain basis vectors and frequency domain basis vectors; transmitting a reference signal to the terminal; receiving quantized transformed precoding vectors based on the reference signal from the terminal, In the same field of endeavor, however, GAO teaches: transmitting, to a terminal, information on spatial domain basis vectors and frequency domain basis vectors; (¶ 0060: [N]etwork device 2 is a device that can communicate with the terminal device 1, and may be a base station ; ¶ 0075: [A] two-dimensional spatial-frequency combination coefficient corresponding to both a spatial domain basis vector and a frequency domain basis vector ; ¶ 0077: The network device sends configuration information to the terminal ; ¶ 0078: [C]onfiguration information may indicate one or more spatial domain beam basis vector groups and Q thresholds ; ¶ 0027: Configuration information indicating one or more spatial domain beam basis vector groups and Q thresholds is sent to a terminal device. Amplitudes and phases of spatial-frequency combination coefficients in M spatial-frequency combination coefficient vectors are received from the terminal device. The Q thresholds correspond one-to-one to spatial domain beam basis vectors in the one or more spatial domain beam basis vector groups. The M spatial-frequency combination coefficient vectors are determined based on L spatial domain beam basis vectors, K frequency domain basis vectors corresponding to each of the L spatial domain beam basis vectors, and a target precoding vector ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s machine learning-based CSI feedback to provide receiving information on spatial domain basis and frequency domain basis vectors as taught by GAO to provide transforming precoding vectors and quantizing the precoding vectors, such that indication information of the precoding vector is reported to the network device, so as to help the network device obtain an optimal precoding vector. See GAO, at ¶ 0069. Also, in the same field of endeavor, however, LIU teaches: transmitting a reference signal to the terminal; (¶ 0007: [R]eceiving, by a terminal, a downlink reference signal sent by an access network device ) receiving quantized transformed precoding vectors based on the reference signal from the terminal; (¶ 0170: Step 404: The terminal sends SRSs to the access network device on a plurality of uplink SRS resources, where SRSs sent on different uplink SRS resources are precoded by using different uplink precoding vectors ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s “offloading”—from the base station to the UE—of the information learned during its offline learning stage to provide for sending/receipt of a reference signal as taught by LIU to obtain precoding vectors such that—based on a channel reciprocity principle—the precoding vector is more suitable for the actual case of the uplink channel of the terminal, so that the transmission performance of the uplink data can be improved. See LIU, at ¶ 0120. NERINI does not explicitly disclose: dequantizing the quantized transformed precoding vectors to generate the transformed precoding vectors; and In the same field of endeavor, however, ZHU teaches: dequantizing the quantized transformed precoding vectors to generate the transformed precoding vectors; and (¶ 0101: [P]rocessing circuitry maps each phase, of the second half of the phases, to the CDF; wherein the processing circuitry evenly quantizes a phase probability in a range between 0 and 1 to form a quantized probability; wherein the processing circuitry maps the quantized probability using an inverse of the CDF to a quantized phase; and wherein the processing circuitry sequentially quantizes phases of each column of the CDF ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s machine learning-based CCSI feedback procedure to provide plots of cumulative distribution function (CDF) of the first 7 Givens rotation angles as taught by ZHU so that the distribution of the Givens angle becomes non-evenly distributed and more concentrated to small values and this small-value concentration may be used to increase the quantization accuracy, such that the small-value concentration may be used to reduce the quantization overhead. See ZHU, at ¶ 0120. The Examiner finds that because the following “wherein” clauses merely reference (1) formation of quantized transformed precoding vectors and (2) generation of transformed precoding vectors, formation/generation operations occurring at the UE or elsewhere generally do not limit the base station method, and thus the clauses are entitled to little or no patentable weight. That is, because claim 15 only requires precoding vectors having some non-descript origin and/or characteristic, how another entity—here, the UE—formed/generated them does not meaningfully limit the base station method. In any event, the applied are discloses or suggests the subject matter of the “wherein” clauses for at least the reasons given for claim 1. Regarding claim 17, the combination of NERINI and LIU, as applied above, renders obvious the method of claim 15. NERINI further discloses: restoring an original precoding vector from the reconstructed at least one precoding vector by using an artificial neural network. (Pg. 2, 3rd ¶: [A] fully connected neural network (NN) is used to map the uplink to the corresponding downlink channels ) Regarding claim 18, the combination of NERINI and LIU, as applied above, renders obvious the method of claim 17. NERINI further discloses: wherein the artificial neural network is one of a fully-connected neural network, a multi-layer perceptron, a convolutional neural network, or a transformer. (Pg. 2, 3rd ¶: The same task is solved more efficiently . . . with a convolutional neural network (CNN) treating the space-frequency channel matrix as an image ) 07-21-aia AIA Claim s 5-7 are rejected under 35 U.S.C. § 103 as being unpatentable over NERINI in view of LIU and GAO, and further in view of US 2016/0149617 (hereinafter, “ZHU”) . Regarding claim 5, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 1. NERINI further discloses: wherein the quantizing of the transformed precoding vectors comprises: transforming elements of the transformed precoding vectors by applying a distribution function; and (§ V.B: In Fig. 4, the NMSE and the cosine similarity ρ between the reconstructed and the true downlink channel matrices are reported. These cumulative distribution functions (CDFs), calculated over the whole test set, show that the reconstruction quality increases with B, and that, more importantly, no performance upper bound is present ) NERINI does not explicitly disclose: quantizing the elements transformed through the distribution function. In the same field of endeavor, however, ZHU teaches: quantizing the elements transformed through the first distribution function. (¶ 0052: [T]he CDF curves in FIG. 2 may quantize each phase into N s bits, as in Eq. (20) ; ¶ 0100: [P]rocessing circuitry quantizes each phase, of the second half of the phases, in the range between 0 and 2*pi using a cumulative distribution function (CDF) ¶ 0101: [P]rocessing circuitry maps each phase, of the second half of the phases, to the CDF; wherein the processing circuitry evenly quantizes a phase probability in a range between 0 and 1 to form a quantized probability; wherein the processing circuitry maps the quantized probability using an inverse of the CDF to a quantized phase; and wherein the processing circuitry sequentially quantizes phases of each column of the CDF ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s machine learning-based CCSI feedback procedure to provide plots of cumulative distribution function (CDF) of the first 7 Givens rotation angles as taught by ZHU so that the distribution of the Givens angle becomes non-evenly distributed and more concentrated to small values and this small-value concentration may be used to increase the quantization accuracy, such that the small-value concentration may be used to reduce the quantization overhead. See ZHU, at ¶ 0120. Regarding claim 6, the combination of NERINI, LIU, GAO, and ZHU, as applied above, renders obvious the method of claim 5. NERINI further discloses: wherein the distribution function is one of a cumulative distribution function (CDF) function, a CDF function approximated as a normal distribution function, or a normal distribution function. (§ V.B: In Fig. 4, the NMSE and the cosine similarity ρ between the reconstructed and the true downlink channel matrices are reported. These cumulative distribution functions (CDFs), calculated over the whole test set, show that the reconstruction quality increases with B, and that, more importantly, no performance upper bound is present ) Regarding claim 7, the combination of NERINI, LIU, GAO, and ZHU, as applied above, renders obvious the method of claim 5. NERINI further discloses: wherein in the quantizing of the elements transformed through the distribution function, the quantizing is performed using at least one of a uniform quantization scheme (§ V.B: [F]eedback bits are allocated uniformly to all the latent space dimensions [in determining quantization levels]) , a scalar quantization scheme (App’x § C: [C]onsider the scaled dataset Y=cX, in which y n is the n-th point ) , or a quantization scheme using a Lloyd Max quantizer . 07-21-aia AIA Claim 4 is rejected under 35 U.S.C. § 103 as being unpatentable over NERINI, LIU, and GAO, as applied above, and further in view of US 2021/0056540 (hereinafter, “MCCAULEY”) . Regarding claim 4, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 1. NERINI does not explicitly disclose: wherein in the generating of the transformed precoding vectors, the transformed precoding vectors are generated by performing dimensionality reduction transformation on the at least one precoding vector through an independent component analysis scheme. In the same field of endeavor, however, MCCAULEY teaches: wherein in the generating of the transformed precoding vectors, the transformed precoding vectors are generated by performing dimensionality reduction transformation on the at least one precoding vector through an independent component analysis scheme. (¶ 0112: [D]imensionality reduction techniques can be used by the feature extraction module to reduce a dimensionality of the feature vector: independent component analysis ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s dimensionality reduction procedure to provide a feature extraction module to reduce a dimensionality of the precoding vector using independent component analysis such that dimensionality reduction procedures reduce redundancy in the data points by transforming the data points into a reduced set of features (the feature vector). See MCCAULEY, at ¶ 0112 . 07-21-aia AIA Claim s 16 and 19 are rejected under 35 U.S.C. § 103 as being unpatentable over NERINI, LIU, and GAO, as applied above, and further in view of RAZINKIN . Regarding claim 16, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 15. NERINI does not explicitly disclose: wherein in the reconstructing of the at least one precoding vector, the at least one precoding vector is reconstructed from the transformed precoding vector in an inverse principal component analysis scheme. In the same field of endeavor, however, RAZINKIN teaches: wherein in the reconstructing of the at least one precoding vector, the at least one precoding vector is reconstructed from the transformed precoding vector in an inverse principal component analysis scheme. (Pg. 8, 2nd ¶: A second aspect of this disclosure provides a device for reconstructing CSI from compressed CSI, the CSI being related to a channel between a transmitter and a receiver, and the device being configured to: calculate a 2D channel matrix based on the compressed CSI . . . [W]herein the channel coefficients for the N transmitting antenna ports and the M transmission layers or M receiving antenna ports are calculated by performing an inverse spatial transformation based on K spatial beam direction vector components, and/or the channel coefficients for the F frequency subbands are calculated by performing a time-to-frequency transformation based on basis vector components selected for the T time delay taps common for the M transmission layers or M receiving antenna ports ; Pg. 9, 1st full ¶: In an implementation form of the second aspect, . . . the inverse spatial transformation comprises . . . a PCA based inverse transformation ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s dimensionality reduction procedure to provide vector reconstruction from the transformed precoding vector using PCA based inverse transformation as taught by RAZINKIN, so as to enable a decompression of the compressed CSI and thereby support the reduction of the overhead without substantially sacrificing accuracy. See RAZINKIN, Pg. 8, 3rd ¶. Regarding claim 19, the combination of NERINI, LIU, and GAO, as applied above, renders obvious the method of claim 15. NERINI further discloses: wherein the quantized transformed precoding vectors are received together with information on a number of reduced dimensions and information on a bit length for each dimension, and (§ V.B: [T]he feedback bits are allocated uniformly to all the latent space dimensions, and the quantization levels are determined with k-means clustering ) NERINI does not explicitly disclose: the transformed precoding vectors are generated through dequantization on the quantized transformed precoding vectors using the information on the number of reduced dimensions and the information on the bit length. In the same field of endeavor, however, RAZANKIN teaches: the transformed precoding vectors are generated through dequantization on the quantized transformed precoding vectors using the information on the number of reduced dimensions and the information on the bit length. (Pg. 8, 4th ¶: [O]btain a number of L quantized and indexed channel coefficients from the compressed CSI; and dequantize and rearrange the L quantized and indexed channel coefficients to obtain the 2D channel matrix ; [T]o reconstruct the CSI from the compressed CSI 203, it is enough firstly to dequantize the coefficients and then, re-use the spatial basis (long term) with size (N*M) x K. and frequency basis (inverse transform to the respective frequency-to-time delay compression stage with length F), and multiply the K x T projected coefficients back on the respective indexed components of each basis (spatial and frequency basis). As a result, a reconstructed channel with dimensions N*M x F is obtained, which can be easily reshaped into a 3D tensor 402 with size N x F x M ) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify NERINI’s dimensionality reduction procedure to provide obtain a number of L quantized and indexed channel coefficients from the compressed CSI as taught by RAZINKIN, so as to dequantize and rearrange the L quantized and indexed channel coefficients to obtain the 2D channel matrix and thereby reconstruct the CSI from the compressed CSI. See RAZINKIN, Pg. 23, 3rd ¶. Conclusion Any inquiry concerning this communication or earlier communications from the Examiner should be directed to Garth D Richmond whose telephone number is (703)756-4559. The Examiner can normally be reached M-F 8 a.m. - 5 p.m. ET. 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, Kathy Wang-Hurst can be reached at 571-270-5371. 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. /GARTH D RICHMOND/Examiner, Art Unit 2644 /KATHY W WANG-HURST/Supervisory Patent Examiner, Art Unit 2644 Application/Control Number: 18/361,279 Page 2 Art Unit: 2644 Application/Control Number: 18/361,279 Page 3 Art Unit: 2644 Application/Control Number: 18/361,279 Page 4 Art Unit: 2644
Prosecution Insights