DETAILED ACTION
This Office Action is in response to the application filed on 02/13/2025.
Claims 4, 6, 21-25 and 28-52 are cancelled.
Claims 1-3, 5, 7-20, and 26-27 are presented for examination.
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 statement (IDS) submitted on 02/13/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 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)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1-3, 8, 11-13, 15, 18, 26-27 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Niu et al. (hereinafter Niu, US 2023/0283340 A1).
Regarding Claim 1, Niu discloses A method of wireless communications at a first network entity associated with a test equipment vendor (Niu: Fig. 5 cellular base station which takes part in testing the model in para.0109, which is associated with an infrastructure vendor in para.0112 “infrastructure vendor”, para.0098 device vendor), the method comprising:
receiving, at the first network entity (Niu: Fig. 5 para.0100 cellular base station), information specifying at least one of a type of machine learning model to use for a machine learning decoder (Niu: para.0101 “According to such a framework, it may be the case that the artificial intelligence encoder and the artificial intelligence decoder for the artificial intelligence model selected by the wireless device are trained via centralized offline training at the wireless device side….The artificial intelligence model (or at least the artificial intelligence decoder portion) may be provided from the wireless device to the cellular base station once the wireless device has selected the artificial intelligence model to use to perform channel state information reporting. ” a particular type of machine learning model, Artificial intelligence decoder, may be provider to the cellular base station),
one or more parameters for the machine learning decoder (Niu: para.0100 “In such a scenario, the wireless device may send the refined artificial intelligence decoder weights to the cellular base station, e.g., for aggregation. At least in some instances, such modifications and fine tuning may be limited to model weights/bias without affecting the structure of the artificial intelligence model. The cellular base station may be able to aggregate the feedback of the wireless device and potentially similar feedback from other wireless devices and update the artificial intelligence model decoder.” The cellular base station receives decoder weights for the AI model decoder.),
or one or more key performance indicators for the machine learning decoder;
receiving, at the first network entity, a representation of control information from a second network entity (Niu: para.0103 “In 508, the wireless device may perform channel state information reporting using the determined artificial intelligence model. The channel state information reporting using the determined artificial intelligence model may be performed for a scheduled periodic channel state information report or for an aperiodic channel state information (AP-CSI) request.” Para.0109 “Thus, in a system using the approach of FIG. 6 to implement AI based CSI reporting, a UE 602 may receive a downlink channel (“H”) and may perform encoder inferencing and provide AI based CSI feedback to a gNB 604. ” Control state information is received by the cellular base station 604 in Fig. 6, and/or the channel state information is received by the cellular base station in step 508 of Fig. 5.); and
processing, using the machine learning decoder configured based on the received information (Niu: para.0100 “In such a scenario, the wireless device may send the refined artificial intelligence decoder weights to the cellular base station, e.g., for aggregation. At least in some instances, such modifications and fine tuning may be limited to model weights/bias without affecting the structure of the artificial intelligence model.” The decoder model configured with the obtained model weights), the representation of the control information to generate a reconstruction of the control information (Niu: para.0109 “The gNB 604 may perform decoder inferencing to determine a reconstruction channel (“Ĥ”). The AI based CSI feedback mechanism may be trained to minimize normalized mean square error (NMSE)” the base station reconstructs the CSI resulting in reconstruction channel Ĥ).
Regarding Claim 2, Niu discloses claim 1 as set forth above.
Niu further discloses wherein the control information comprises channel state information (CSI) or channel state feedback (CSF) (Niu: para.0109 “Thus, in a system using the approach of FIG. 6 to implement AI based CSI reporting, a UE 602 may receive a downlink channel (“H”) and may perform encoder inferencing and provide AI based CSI feedback to a gNB 604. The gNB 604 may perform decoder inferencing to determine a reconstruction channel (“Ĥ”). The AI based CSI feedback mechanism may be trained to minimize normalized mean square error (NMSE), which may be defined using the following equation, at least according to some embodiments.” CSI feedback information is the control information, as channel state feedback is based on CSI, the channel state feedback of Niu reads on both CSI and CSF).
Regarding Claim 3, Niu discloses claim 1 as set forth above.
Niu further discloses wherein the information specifies the type of machine learning model to use for the machine learning decoder (Niu: para.0101 “According to such a framework, it may be the case that the artificial intelligence encoder and the artificial intelligence decoder for the artificial intelligence model selected by the wireless device are trained via centralized offline training at the wireless device side….The artificial intelligence model (or at least the artificial intelligence decoder portion) may be provided from the wireless device to the cellular base station once the wireless device has selected the artificial intelligence model to use to perform channel state information reporting. ” a particular type of machine learning model, Artificial intelligence decoder, may be provider to the cellular base station), and
wherein the information further specifies the one or more parameters for the machine learning decoder (Niu: para.0100 “In such a scenario, the wireless device may send the refined artificial intelligence decoder weights to the cellular base station, e.g., for aggregation. At least in some instances, such modifications and fine tuning may be limited to model weights/bias without affecting the structure of the artificial intelligence model. The cellular base station may be able to aggregate the feedback of the wireless device and potentially similar feedback from other wireless devices and update the artificial intelligence model decoder.” The cellular base station receives decoder weights for the AI model decoder.).
Regarding Claim 8, Niu discloses claim 1 as set forth above.
Niu further discloses further comprising: determining, based on the reconstruction of the control information, at least one of a precoding matrix or a rank of one or more antennas of the first network entity (Niu: para.0074 “As a detailed example, in the 3GPP NR cellular communication standard, the channel state information fed back from the UE based on CSI-RS for CSI acquisition may include one or more of a channel quality indicator (CQI), a precoding matrix indicator (PMI), a rank indicator (RI), a CSI-RS Resource Indicator (CRI), a SSBRI (SS/PBCH Resource Block Indicator, and a Layer Indicator (LI), at least according to some embodiments.” para.0109 “The gNB 604 may perform decoder inferencing to determine a reconstruction channel (“Ĥ”). The AI based CSI feedback mechanism may be trained to minimize normalized mean square error (NMSE)” The CSI may include Precoding matrix indicator, which is reconstructed by the decoder at the base station in Fig. 6.).
Regarding Claim 11, Niu discloses claim 1 as set forth above.
Niu further discloses further comprising: configuring, at the first network entity, the machine learning decoder based on the information (Niu: para.0100 “In such a scenario, the wireless device may send the refined artificial intelligence decoder weights to the cellular base station, e.g., for aggregation. At least in some instances, such modifications and fine tuning may be limited to model weights/bias without affecting the structure of the artificial intelligence model. The cellular base station may be able to aggregate the feedback of the wireless device and potentially similar feedback from other wireless devices and update the artificial intelligence model decoder.” The AI decoder is updated using weights and feedback from various devices, i.e. configured).
Regarding Claim 12, Niu discloses claim 1 as set forth above.
Niu further discloses training the machine learning decoder using data based on a set of profiles specified for the data (Niu: para.0101 “Such training for one or more models may be performed by a wireless device vendor using aggregated field collected data and/or simulation data for various cell types, various channel conditions and/or other scenario elements, and the trained model(s) may be provided to wireless devices associated with that wireless device vendor, as one possibility. As another possibility, the training may be performed by the wireless device using field collected data from the wireless device.” Para.0102 “According to such a framework, artificial intelligence model training may be performed on the wireless device side for the artificial intelligence encoder, and on the network side for the artificial intelligence decoder. ” The machine learning decoder is trained using data that is based on a set of profiles for the data, i.e. data of various cell types and channel conditions.).
Regarding Claim 13, Niu discloses claim 12 as set forth above.
Niu further discloses wherein the set of profiles for the data comprises one or more parameters associated with at least one of a propagation channel condition (Niu: para.0101 “Such training for one or more models may be performed by a wireless device vendor using aggregated field collected data and/or simulation data for various cell types, various channel conditions and/or other scenario elements,” channel conditions), an antenna configuration for the first network entity, or a device type.
Regarding Claim 15, Niu discloses claim 1 as set forth above.
Niu further discloses wherein the information specifies a single type of machine learning model to use for the machine learning decoder for all profiles in the set of profiles (Niu: para.0101 “According to such a framework, it may be the case that the artificial intelligence encoder and the artificial intelligence decoder for the artificial intelligence model selected by the wireless device are trained via centralized offline training at the wireless device side….The artificial intelligence model (or at least the artificial intelligence decoder portion) may be provided from the wireless device to the cellular base station once the wireless device has selected the artificial intelligence model to use to perform channel state information reporting. ” a particular type of machine learning model, Artificial intelligence decoder, may be provider to the cellular base station, thereby specifying a single possible machine learning model to use regardless of profile, i.e. network condition device type etc as mentioned in claim 13.).
Regarding Claim 18, Niu discloses claim 1 as set forth above.
Niu further discloses wherein a machine learning encoder of the second network entity is trained using data generated based on the machine learning decoder of the first network entity (Niu: para.0100 “The cellular base station may further provide an update to the wireless device to indicate whether the artificial intelligence model decoder has been updated, which may affect which potential modifications (if any) to the artificial intelligence model encoder to use when performing channel state information reporting using the determined artificial intelligence model, for example in order to maintain joint optimization of the encoder and decoder portions of the artificial intelligence model.”).
Regarding Claims 26-27, they teach all of the same elements as claims 1-2 but in A first network entity associated with a test equipment vendor (Niu: Fig. 5 cellular base station which takes part in testing the model in para.0109, which is associated with an infrastructure vendor in para.0112 “ infrastructure vendor”), the first network entity comprising: at least one memory; and at least one processor coupled to the at least one memory and configured to (Niu: para.0043). Therefore the supporting rationale for the rejection to claims 1-2 applies equally as well to that of claims 26-27.
Claim Rejections - 35 USC § 103
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) 5, 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Niu et al. (hereinafter Niu, US 2023/0283340 A1) in view of Lindbom et al. (hereinafter Lind, US 2024/0313839 A1).
Regarding Claim 5, Niu discloses claim 1 as set forth above.
Niu further discloses wherein the information specifies the type of machine learning model to use for the machine learning decoder (Niu: para.0101 “According to such a framework, it may be the case that the artificial intelligence encoder and the artificial intelligence decoder for the artificial intelligence model selected by the wireless device are trained via centralized offline training at the wireless device side….The artificial intelligence model (or at least the artificial intelligence decoder portion) may be provided from the wireless device to the cellular base station once the wireless device has selected the artificial intelligence model to use to perform channel state information reporting. ” a particular type of machine learning model, Artificial intelligence decoder, may be provider to the cellular base station).
However Niu does not explicitly disclose and the one or more key performance indicators for the machine learning decoder, wherein the information specifies the one or more key performance indicators for the machine learning decoder.
Lind discloses wherein the information specifies the one or more key performance indicators for the machine learning decoder, wherein the information specifies the one or more key performance indicators for the machine learning decoder (Lind: para.0087 “In some examples, the resulting performance loss may be compared with a threshold value. In case, the resulting performance loss is higher than the threshold value, then the CSI compression quality is classified with label “1” which indicates that the performance loss has a larger impact on application performance. When the resulting performance loss is lower than the threshold value, then the CSI compression quality is classified with label “0” which indicates that the performance loss has no or minor impact on the application performance. Thus, the neural network based classifier is used for classification of the CSI compression quality.” Para.0088 “In this case, the UE has implemented a (reference) decoder that is used to reconstruct the compressed CSI. The threshold value for comparing the resulting performance loss for classification of the CSI compression quality may in such cases be indicated to the UE by the base station.” The entity comprising the decoder is provided with performance thresholds for evaluating reconstruction).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Lind in order to incorporate wherein the information specifies the one or more key performance indicators for the machine learning decoder, wherein the information specifies the one or more key performance indicators for the machine learning decoder.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved decoder performance by evaluating reconstruction of the signal (Lind: para.0088)
Regarding Claim 7, Niu discloses claim 1 as set forth above.
However Niu does not explicitly disclose determining, at the first network entity, a quality of the reconstruction of the control information based on the one or more key performance indicators.
Lind discloses determining, at the first network entity, a quality of the reconstruction of the control information based on the one or more key performance indicators (Lind: para.0087 “In some examples, the resulting performance loss may be compared with a threshold value. In case, the resulting performance loss is higher than the threshold value, then the CSI compression quality is classified with label “1” which indicates that the performance loss has a larger impact on application performance. When the resulting performance loss is lower than the threshold value, then the CSI compression quality is classified with label “0” which indicates that the performance loss has no or minor impact on the application performance. Thus, the neural network based classifier is used for classification of the CSI compression quality.” Para.0088 “In this case, the UE has implemented a (reference) decoder that is used to reconstruct the compressed CSI. The threshold value for comparing the resulting performance loss for classification of the CSI compression quality may in such cases be indicated to the UE by the base station.” The entity comprising the decoder is provided with performance thresholds for evaluating reconstruction, and the UE (the decoder side), determines the performance loss based on the kpi).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Lind in order to incorporate determining, at the first network entity, a quality of the reconstruction of the control information based on the one or more key performance indicators.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved decoder performance by evaluating reconstruction of the signal (Lind: para.0088)
Claim(s) 9-10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Niu et al. (hereinafter Niu, US 2023/0283340 A1) in view of Chen et al. (hereinafter Chen, US 2014/0133317 A1).
Regarding Claim 9, Niu discloses claim 8 as set forth above.
However Niu does not explicitly disclose determining, at the first network entity, a performance quality of the second network entity based on a comparison of least one of the precoding matrix or the rank to at least one of a reference precoding matrix or a reference rank.
Chen discloses determining, at the first network entity, a performance quality of the second network entity based on a comparison of least one of the precoding matrix or the rank to at least one of a reference precoding matrix or a reference rank (Chen: para.0112 “Whereas in closed loop MIMO there are at least three constellations corresponding to different PMIs that must be decided by the wireless terminal, the technology disclosed herein by using open loop MIMO has the base station node look at all constellation options and chooses the best PMI that the node can perceive, which is essentially the equivalent of a closed loop MIMO technique, without the overhead of closed loop MIMO (which involved the wireless terminal sending the PMI(s), the CQIs of two codewords, etc.). In so doing, the transmit switch controller 40 calculates the throughput value, and continually checks whether the throughput value based on a first PMI (PMI.sub.1) or a second PMI (PMI.sub.2) is better.” The network entity compares the throughput from each PMI, i.e. the determined PMI and a reference PMI, and determines which has a higher throughput, i.e. determining a throughput gain based on the comparison.).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Chen in order to incorporate determining, at the first network entity, a performance quality of the second network entity based on a comparison of least one of the precoding matrix or the rank to at least one of a reference precoding matrix or a reference rank.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved throughput of the network (Chen: para.0112).
Regarding Claim 10, Niu-Chen disclose claim 9 as set forth above.
However Niu does not explicitly disclose wherein the performance quality is based on a throughput gain.
Chen discloses wherein the performance quality is based on a throughput gain (Chen: para.0112 “Whereas in closed loop MIMO there are at least three constellations corresponding to different PMIs that must be decided by the wireless terminal, the technology disclosed herein by using open loop MIMO has the base station node look at all constellation options and chooses the best PMI that the node can perceive, which is essentially the equivalent of a closed loop MIMO technique, without the overhead of closed loop MIMO (which involved the wireless terminal sending the PMI(s), the CQIs of two codewords, etc.). In so doing, the transmit switch controller 40 calculates the throughput value, and continually checks whether the throughput value based on a first PMI (PMI.sub.1) or a second PMI (PMI.sub.2) is better.” The network entity compares the throughput from each PMI, i.e. the determined PMI and a reference PMI, and determines which has a higher throughput, i.e. determining a throughput gain based on the comparison.).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Chen in order to incorporate wherein the performance quality is based on a throughput gain.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved throughput of the network (Chen: para.0112).
Claim(s) 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Niu et al. (hereinafter Niu, US 2023/0283340 A1) in view of Selokar et al. (hereinafter Selokar, US 2021/0273843 A1).
Regarding Claim 14, Niu discloses claim 12 as set forth above.
Niu further discloses the set of data being provided by a respective vendor (Niu: para.0101 “Such training for one or more models may be performed by a wireless device vendor using aggregated field collected data and/or simulation data for various cell types, various channel conditions and/or other scenario elements, and the trained model(s) may be provided to wireless devices associated with that wireless device vendor, as one possibility.” The vendor may perform the training therefore the dataset used to train the machine learning model is from the vendor)
However while Niu implies the use of information from a plurality of vendors in para.0090, and para.0117, Niu does not explicitly disclose wherein the data is comprised of multiple sets of data from a plurality of vendors, each set of data of the multiple sets of data being provided by a respective vendor of the plurality of vendors.
Selokar discloses wherein the data is comprised of multiple sets of data from a plurality of vendors, each set of data of the multiple sets of data being provided by a respective vendor of the plurality of vendors (Selokar: para.0018 “To address such challenge, current invention provides a system and method for recognizing and addressing network alarms in a computer network. Specifically, a trained data model is used to recognize and address the network alarms arising in the computer network. The trained data model are developed by learning upon multiple network alarms, from multiple devices of multiple vendors, indicating operating conditions of different network devices, and attributes pre-identified to be associated with each of the multiple network alarms.” Training data from a plurality of vendors are used in combination to train the device).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Selokar in order to incorporate wherein the data is comprised of multiple sets of data from a plurality of vendors, each set of data of the multiple sets of data being provided by a respective vendor of the plurality of vendors.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved model performance by using data form a wider array of sources (Selokar), which is a concept that is briefly addressed in Niu para.0090 for the model to be applicable in cells associated with different vendors (Niu: para.0090).
Claim(s) 16-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Niu et al. (hereinafter Niu, US 2023/0283340 A1) in view of Yelahanka Raghuprasad et al. (hereinafter Yel, US 2021/0281491 A1).
Regarding Claim 16, Niu discloses Claim 12 as set forth above.
However Niu does not explicitly disclose wherein the information specifies a first type of machine learning model to use for the machine learning decoder for at least a first profile in the set of profiles and a second type of machine learning model to use for the machine learning decoder for at least a second profile in the set of profiles.
Yel disclose wherein the information specifies a first type of machine learning model to use for the machine learning decoder for at least a first profile in the set of profiles and a second type of machine learning model to use for the machine learning decoder for at least a second profile in the set of profiles (Yel: para.0065 “In some embodiments, NACO 306 may also provide guidelines to the edge devices so that they can choose which compression model to use, given the performance vs. compression tradeoff. In one embodiment, NACO 306 might send simple rules on which model to use for a given timeframe. For example, as shown, NACO 306 may send model selection rules 312 to edge compressor 502, to select which compression model 402 should be used in a given timeframe. In this case, NACO 306 might examine the usage of network bandwidth over a given time-period and decide to utilize different models 402 at different times-of-the-day/days-of-the-week. For example, NACO 306 might determine that edge compressor 502 should use model M1 (e.g., less aggressive compression) during most crucial, but less congested, times of the day. However, NACO 306 may determine that edge compressor 502 should switch to model M2 (more aggressive compression) on weekends and nights, times at which it is generally expected a lower congestion.” Information can be provided to the edge devices to select particular compression models, seen in Fig. 4 compression models include decoders para.0053, for certain profiles, i.e. based on different channel conditions by timeframe.).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Yel in order to incorporate wherein the information specifies a first type of machine learning model to use for the machine learning decoder for at least a first profile in the set of profiles and a second type of machine learning model to use for the machine learning decoder for at least a second profile in the set of profiles.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of optimal network performance based on network conditions (Yel: para.0065).
Regarding Claim 17, Niu discloses claim 12 as set forth above.
However Niu does not explicitly disclose wherein the information specifies a separate type of machine learning model to use for the machine learning decoder for each profile in the set of profiles.
Yel disclose wherein the information specifies a separate type of machine learning model to use for the machine learning decoder for each profile in the set of profiles (Yel: para.0065 “In some embodiments, NACO 306 may also provide guidelines to the edge devices so that they can choose which compression model to use, given the performance vs. compression tradeoff. In one embodiment, NACO 306 might send simple rules on which model to use for a given timeframe. For example, as shown, NACO 306 may send model selection rules 312 to edge compressor 502, to select which compression model 402 should be used in a given timeframe. In this case, NACO 306 might examine the usage of network bandwidth over a given time-period and decide to utilize different models 402 at different times-of-the-day/days-of-the-week. For example, NACO 306 might determine that edge compressor 502 should use model M1 (e.g., less aggressive compression) during most crucial, but less congested, times of the day. However, NACO 306 may determine that edge compressor 502 should switch to model M2 (more aggressive compression) on weekends and nights, times at which it is generally expected a lower congestion.” Information can be provided to the edge devices to select particular compression models, seen in Fig. 4 compression models include decoders para.0053, for each profiles, i.e. based on different channel conditions by timeframe.).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Yel in order to incorporate wherein the information specifies a separate type of machine learning model to use for the machine learning decoder for each profile in the set of profiles.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of optimal network performance based on network conditions (Yel: para.0065).
Claim(s) 19-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Niu et al. (hereinafter Niu, US 2023/0283340 A1) in view of Tan et al. (hereinafter Tan, US 2023/0354096 A1).
Regarding Claim 19, Niu discloses claim 1 as set forth above.
However Niu does not explicitly disclose wherein the representation of the control information is a latent representation of the control information.
Tan discloses wherein the representation of the control information is a latent representation of the control information (Tan: para.0039 “During the training phase, the encoder 202 may take at least one input x 210A, which in this example is a data sample of CSI, such as channel gain.” para.0040 “The output 210B of the encoder 202 may be modelled as a log-likelihood ratio (LLR) vector λ 210B, which is a log-likelihood ratio (LLR) value of a latent variable bit z.sub.j. The latent variable represents a lower dimension hidden layer into which the encoder encodes the data sample 210A. In this way, the encoder can provide compression by encoding the input into a lower dimensional latent domain.” The LLR from the encoder that is sent to the decoder for reconstruction is a latent representation of CSI).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Tan in order to incorporate wherein the representation of the control information is a latent representation of the control information.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved bandwidth and improved processing by using a lower dimension version of the data (Tan: para.0040).
Regarding Claim 20, Niu-Tan discloses claim 19 as set forth above.
However Niu does not explicitly disclose wherein the latent representation of the control information comprises a feature vector representing the control information.
Tan disclose wherein the latent representation of the control information comprises a feature vector representing the control information (Tan: para.0039 “During the training phase, the encoder 202 may take at least one input x 210A, which in this example is a data sample of CSI, such as channel gain.” para.0040 “The output 210B of the encoder 202 may be modelled as a log-likelihood ratio (LLR) vector λ 210B, which is a log-likelihood ratio (LLR) value of a latent variable bit z.sub.j. The latent variable represents a lower dimension hidden layer into which the encoder encodes the data sample 210A. In this way, the encoder can provide compression by encoding the input into a lower dimensional latent domain.” Para.0041 “ In other words, the binary sampler output 212A is a binary CSI coding sequence (which represents the encoded CSI code) with each bit z.sub.j ∈ {0, 1}.” the output is a binary vector representation of the CSI, i.e. features.).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Niu with Tan in order to incorporate wherein the latent representation of the control information comprises a feature vector representing the control information.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved bandwidth and improved processing by using a lower dimension version of the data (Tan: para.0040).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. O’Shea et al. US 2018/0367192 A1, see para.0095 and Fig. 1 showing decoder reconstruction of CSI.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to EUI H KIM whose telephone number is (571)272-8133. The examiner can normally be reached 7:30-5 M-R, M-F alternating.
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, Kamal B Divecha can be reached at 5712725863. 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.
/EUI H KIM/ Examiner, Art Unit 2453
/KAMAL B DIVECHA/ Supervisory Patent Examiner, Art Unit 2453
Prosecution Insights