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 .
Claims 1-5, 7, 9-14, 17, 18-24 are pending.
Claims 12-14, 17, 18, and 20 are withdrawn.
Response to Arguments
Arguments presented in Remarks 11/25/2025 are fully considered but they are not persuasive.
In Remarks pages 13-14, Applicant provides their own reading of reference of record Gunduz and a narration of the invention as intended in the Specification. The examiner asserts that, while the expertise provided is appreciated, the interpretation of claims and references are subjected to MPEP guidance.
In page 16, Applicant argues the input layer 102 and output layer 112 belong to different seprate individual neural network models while the claimed invention is directed to the same network model.
The examiner respectfully disagrees. Per ¶0144, it is explicitly clearly stated that 102 and 112 are parts of a neural network 100, which has an encoder 202 and a decoder 204. Gunduz further in ¶0149 states “the encoder and decoder neural networks can be used as a form of joint source channel coding. This provides simplicity over having two separate systems for encoding and two separate systems for decoding”. Thus it is undisputable that the layers in questions are part of a single neural network that has multiple sub-networks.
Applicant further argues :
In Gunduz, the encoder neural network 202 (e.g., equivalent to the first communication device of amended claim 1 of the present application) includes input layers 102, 106 and 110 comprising a plurality of nodes 104, 108 and 112. That is, 108 represents the nodes comprised in layer 106. Therefore, layer 106 and layer 110 belong to different layers in the encoder neural network 202, not different sub-modules in the same layer.
However, in amended claim 1, the current input information of the target sub- module of the first artificial intelligence network model includes at least one of the following: current input information of another sub-module of a same level as the target sub-module; input information at a previous moment of another sub-module of a same level as the target sub-
module; current intermediate information of another sub-module of a same level as the target sub-module; intermediate information at a previous moment of another sub-module of a same level as the target sub-module; current output information of another sub-module of a same level as the target sub-module; or output information at a previous moment of another sub-module of a same level as the target sub-module. Therefore, the current input information of the target sub- module of the first artificial intelligence network model includes the related information of another sub-module of the same level as the target sub-module. It is clear that, the target sub- module and another sub-module are different sub-modules in the same level.
The examiner respectfully disagrees. The terminology “level” in the claim is ambiguous, and there is no definitive definition. As seen in neural network which can be divided into encoder level 202 and decoder level 202, it can be seen that 106 and 110 are different layers (i.e. sub-modules) of the same level, and indeed 110 receives inputs data from the outputs of 106. The argument is not persuasive.
In pages 18-19, Applicant provides their own interpretation of reference Tian, arguing that Tian has different purposes for dividing first communication information into one or more sub-band information is to obtain signal-to-noise ratio information, as such is not analogous or not bodily incorporation.
The examiner respectfully disagrees. Dividing information input into different sub-band pieces are well-established in information processing field of endeavor, regardless of individual purposes. This practice mainly is to improve multi-resolution analysis and efficiency via parallel processing. Even if granted Tian has a different intended use, but the principle for such pre-processing of inputs are the same in many types of systems. Both references Gunduz and Tian concern with solving problem/improvement of exchanging information using artificial AI network model. As such, PHOSITA is not limited to intended purpose in each references but rather is encouraged to incorporate Tian’s division of data into sub-band pieces to reap the common benefits stated above.
Applicant further alleges Tian does not disclose the entire neural network structure as claimed in claim 1. The examiner respectfully points that Tian is incorporated for the pre-processing of input step, as a secondary reference. Tian is not required to mirror the exact same limitations that Gunduz already did. There is no such requirement in the MPEP. Tian is a reference of analogous art and/or in the same field of endeavor, and that is sufficient. It is well-established that a secondary reference need not to disclose every element of the claim, provided that combination of the cited references would have suggested the claimed limitations to one of ordinary skill in the art (See In re Keller, 642 F.2d 413).
Therefore, in view of the discussions above, the arguments are not persuasive.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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.
Claim(s) 1-5, 7, 19, 21-23 is/are rejected under 35 U.S.C. 103 as being unpatentable over Gunduz (WO 2020/035683) in view of Tian et al. (CN 111614439) – IDS entries.
As to claims 1 and 19:
Gunduz discloses:
A communication information sending method, and a communication device, being a first communication device, and comprising a processor, a memory, and a program or an instruction that is stored in the memory and that is executable on the processor, wherein the program or the instruction, when executed by the processor, (See abstract, ¶0258, a process, a computer communication device with processor/memory) causes the communication device to perform said method comprising: dividing, by a first communication device, first communication information into one or more pieces (of sub-band information); and inputting broadband information of the first communication information and/or the one or more pieces of sub-band information to a first artificial intelligence network model, and sending second communication information output by the first artificial intelligence network model;
(See Abstract, Fig. 1-3, ¶0218, using neural network to process information from an information source module prior sending the output information. Input from information source 302 is split into various portions for each nodes)
wherein the first artificial intelligence network model comprises at least one level of sub- modules, and each level comprises one or more sub-modules; (¶0144, neural network 100 in Fig. 1 having at least one level 202 and/or 204, each comprising one or more sub-modules, for example the first level 204 can include a first level submodule include submodule 120, 124)
some sub-modules use a same artificial intelligence network structure and/or use a same artificial intelligence network parameter; (Fig. 1, For example, sub-module 110 having a similar network structure as sub-module 118, namely both having 4 independent nodes in parallel arrangement. Similarly, sub-modules 106 vs. 122 having similar structure, i.e. same number of nodes in parallel arrangement)
wherein current input information of a target sub-module of the first artificial intelligence network model comprises at least one of:
current input information of another sub-module of a same level as the target sub- module;
input information at a previous moment of another sub-module of a same level as the target sub-module;
current intermediate information of another sub-module of a same level as the target sub-module;
intermediate information at a previous moment of another sub-module of a same level as the target sub-module;
current output information of another sub-module of a same level as the target sub-module; or
output information at a previous moment of another sub-module of a same level as the target sub-module;
(See Fig. 1, ¶0144, Input data fed into sub-module 110 are the output of sub-module 108 which is co-located in level 202)
wherein the first artificial intelligence network model comprises a first-level sub- module and a second-level sub-module, (Fig. 1, second-level sub-module 102 and first-level sub-module 122 as an example)
wherein the first-level sub-module comprises one or more first sub-modules (Fig. 1, sub-module layer 122 comprises a plurality of nodes, each node themselves is a first sub-modules) , the second-level sub-module is located at a previous level of the first- level sub-module, (Fig. 1, second-level sub-module 102 is located at a previous level (encoder) with respect to 122) and the second-level sub-module comprises N second sub-modules, wherein N is an amount of broadband information and/or a quantity of sub-bands that are input into the first artificial intelligence network model. (¶0218, “an input layer 102 having input nodes 104 corresponding to a sequence of source symbols = {Si, S2,... , S.sub.m} (…) map sequences of source symbols received from the information source 302 directly to a representation as a channel input vector Xj, usable to drive a transmitter 404 to transmit a corresponding signal over a communications channel ”, explicitly disclosing the sub-module 102 having M number of nodes corresponding to M numbers of symbols received from the information source 302)
Gunduz discloses all limitations above, wherein communication information input is distributed into each input nodes, however does not explicitly divide the input information into one or more pieces of sub-band information. (emphasis added).
Tian, in a related field of endeavor, discloses in page 8 and Fig. 4 discloses communication information input is distributed into each input nodes, however does not explicitly divide the input information into one or more pieces of sub-band information (“signal conversion diagram, a single information unit length is N, can orderly take the information unit of the first bit of the information unit with the same length of the sub-carrier wave number in each OFDM sub-carrier, the rest of the information unit is also processed. so as to ensure that all information of each information unit is on one OFDM sub-carrier. In the figure, the A1-DN before conversion represents the serial polarization code sequence, after conversion, A1-AN, B1-BN, C1-CN, D1-DN are respectively on different sub-carriers, traditional serial-to-parallel conversion mode, an information unit is dispersed on a plurality of OFDM sub-carrier for transmission, and the embodiment of the invention provided by the signal conversion method shown in FIG. 4 can ensure that one information unit on the same OFDM sub-carrier transmission ”).
It would have been obvious to one of ordinary skill in the art before the effective filing time of the invention that Gunduz’s system to divide the input information into one or more pieces of sub-band information. Both references Gunduz and Tian concern with solving problem/improvement of exchanging information using artificial AI network model. Gunduz discloses in ¶0218 that transmitted information are mapped into the network model. Tian discloses such mapping can be done via dividing input into parts and on different subcarrier to improve performance and reliability of the transmitter (Page 2 of Tian).
As to claim 2:
Gunduz in view of Tian discloses all limitations of claim 1, wherein the dividing, by a first communication device, first communication information into one or more pieces of sub-band information comprises: dividing the first communication information into one or more pieces of sub-band information according to a target resource of the first communication information, wherein the target resource comprises at least one of: a frequency domain resource, a time domain resource, a spatial domain resource, or a code domain resource. (Tian, Page 8, Fig. 4, “all information of each information unit is on one OFDM sub-carrier. In the figure, the A1-DN before conversion represents the serial polarization code sequence, after conversion, A1-AN, B1-BN, C1-CN, D1-DN are respectively on different sub-carriers, traditional serial-to-parallel conversion mode, an information unit is dispersed on a plurality of OFDM sub-carrier for transmission”)
As to claim 3:
Gunduz in view of Tian discloses all limitations of claim 2, wherein the dividing, by a first communication device, first communication information into one or more pieces of sub-band information comprises at least one of:
dividing the first communication information into one or more pieces of sub-band information on the frequency domain resource by using a frequency domain unit resource as a unit, wherein the frequency domain unit resource comprises at least one of: a resource block (RB), a physical resource block (PRB), a sub-band, a precoding resource block group (PRG), or a bandwidth part (BWP);
dividing the first communication information into one or more pieces of sub-band information on the time domain resource by using a time domain unit resource as a unit, wherein the time domain unit resource comprises at least one of: a subcarrier, an orthogonal frequency division multiplexing (OFDM) symbol, a slot, or a half-slot;
dividing the first communication information into one or more pieces of sub-band information on the spatial domain resource by using a spatial domain unit resource as a unit, wherein the spatial domain unit resource comprises at least one of: an antenna, an antenna element, an antenna panel, a sending/receiving unit, a beam, a layer, a rank, or an antenna angle; or
dividing the first communication information into one or more pieces of sub-band information on the code domain resource by using a code domain unit resource as a unit, wherein the code domain unit resource comprises at least one of: an orthogonal code, a quasi- orthogonal code, or a semi-orthogonal code.
(See Tian discloses in Fig. 4, page 7, Tian disclose dividing the first communication information into one or more pieces of sub-band information on the frequency domain resource by using a frequency domain unit resource as a unit, wherein the frequency domain unit resource comprises at least one of: a resource block (RB), a physical resource block (PRB), a sub-band, a precoding resource block group (PRG), or a bandwidth part (BWP), “all information of each information unit is on one OFDM sub-carrier. In the figure, the A1-DN before conversion represents the serial polarization code sequence, after conversion, A1-AN, B1-BN, C1-CN, D1-DN are respectively on different sub-carriers, traditional serial-to-parallel conversion mode, an information unit is dispersed on a plurality of OFDM sub-carrier for transmission”).
As to claims 4, 21:
Gunduz in view of Tian discloses all limitations of claim 1/19, wherein the first artificial intelligence network model comprises at least one level of sub modules, and each level comprises one or more sub-modules; wherein a sub-module of the one or more sub-modules comprises at least one of: a fully-connected neural network module; a convolutional neural network module; a recurrent neural network module; a residual neural network module; or a preset algorithm module.
(Gunduz, Fig. 1, ¶0144, the neural network 100, including one level 202 and one level 204, each of which includes at least one sub-modules (102, 122), wherein the sub-modules of 202 is neural network implementing a preset algorithm. Page 7 of Tian, neural network being convolutional neural network)
As to claims 5, 22:
Gunduz in view of Tian discloses all limitations of claim 4/21, wherein some sub- modules use a same artificial intelligence network structure and/or use a same artificial intelligence network parameter; wherein the artificial intelligence network structure used by the sub-module is determined by at least one of: an artificial intelligence network type; a combination manner of multiple comprised sub-networks; a quantity of hidden layers; a connection manner between an input layer and a hidden layer; a connection manner between multiple hidden layers; a connection manner between a hidden layer and an input layer; or a quantity of neurons at each layer. (Gunduz, Fig. 1, For example, sub-module 110 having a similar network structure as sub-module 118, namely both having 4 independent nodes in parallel arrangement. Similarly, sub-modules 106 vs. 122 having similar structure, i.e. same number of nodes in parallel arrangement)
As to claims 7, 23:
Gunduz in view of Tian discloses all limitations of claim 4/19, wherein the second communication information comprises output information of a last-level sub-module of the first artificial intelligence network model, or a combination of output information of multiple last- level sub-modules of the first artificial intelligence network model. (See Gunduz, Fig. 1-3, ¶0144, the output of the neural network 100 is the output of sub-module 122 of 204).
Claim(s) 9, 10, and 24 is/are rejected under 35 U.S.C. 103 as being unpatentable over Gunduz (WO 2020/035683) in view of Tian et al. (CN 111614439) and in further view of Jayaweera et al. (US 2020/0153535).
As to claims 9, 24:
Gunduz in view of Tian discloses all limitations of claim 1/19, wherein the first-level sub-module comprises multiple first sub-modules, , and another first sub-module in the multiple first sub-modules represents sub-band information, wherein the another first sub- module is some or all of the multiple first sub-modules except the at least one first sub-module. (See Fig. 1, first level sub-module 122 comprises N numbers of nodes (neuron module) 124, as established in claim 1, Tian, discloses in page 8 and Fig. 4, at least some nodes processes sub band information).
Neither Tian or Gunduz discloses at least one first sub-module in the first level sub-module represent bandwidth information.
Jayaweera, in a related field of endeavor, discloses in at least ¶0149, wherein communication information (RF signal) is processed by an ANN, wherein aspects of the signal such as frequency (sub band information) and bandwidth information are distributed to respective neuron of the ANN.
It would have been obvious to one of ordinary skill in the art before the effective filing time of the invention that various aspect of the signal input in Gunduz/Tian such as bandwidth to be processed by at least one neuron in 122. Given the goal of Gunduz includes solving problem in reconstruction/reconstitute original information signal, devoting at least one neuron for various respective aspect of a signal (bandwidth for example) allows the system to have full picture of the signal for better reconstruction.
As to claim 10:
Gunduz in view of Tian discloses all limitations of claim 9, wherein input information of the at least one first sub-module is output information of some or all of the second sub-modules, and input information of the another first sub-module is output information of some or all of the second sub-modules. (See Gunduz, Fig. 1, neuron S1 (124) receives input from neuron 120, which also receive inputs from the output of 116)
Allowable Subject Matter
Claim(s) 11 is/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 any intervening claims.
The references of record disclose all limitations of the base claim, however do not disclose: the first artificial intelligence network model further comprises a third-level sub-module, and the third-level sub-module is located at a previous level of the second-level sub-module; wherein the third-level sub-module comprises one third sub-module, and input information of the third sub-module is the broadband information of the first communication information and/or the one or more pieces of sub-band information; wherein input information of one second sub-module is all output information of the third sub-module; or input information of one second sub-module is some output information of the third sub-module, and different second sub-modules has different input information.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 2019/0187718 Systems and methods are provided for detecting features from multi-modal image-like data representations. The system includes a wavelet transformer configured to, via at least one processor, receive image data and to wavelet transform the image data, thereby providing decomposed image data divided into frequency sub-bands. The system further includes an artificial neural network configured to receive and process at least one sub-band of the decomposed image data to detect image features based thereon, the artificial neural network configured to output the detected image features.
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 QUAN M HUA whose telephone number is (571)270-7232. The examiner can normally be reached 10:30-6:30.
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, Anthony Addy can be reached at 571-272-7795. 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.
/QUAN M HUA/Primary Examiner, Art Unit 2645