DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are:
A device configured to: adapt parameters in claim 19.
at least one device configured to adapt parameters in claim 20.
A device configured to determine parameters in claim 22.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112: Indefiniteness
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.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 20 and 21 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 20, the claim recites the limitation said at least one first device. There is insufficient antecedent basis for this limitation in the claim because the term “said at least one first device” lacks antecedent basis. For the purposes of examination, the “said at least one device” is interpreted as the at least one device mentioned at the beginning of the claim.
Regarding claim 21, the claim is rejected for at least its dependence on claim 20.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-22 are rejected under 35 U.S.C 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding claim 1, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A method for adapting parameters of a first neural network. The claim recites a method. A method is one of the four statutory categories of invention.
In Step 2A, Prong 1 of the 101 analysis set forth in MPEP 2106, the examiner has determined that the following limitations recite a process that, under broadest reasonable interpretation, covers a mental process or mathematical concept but for the recitation of generic computer components:
said method comprising: following a detection of a deterioration in the quality of the processing due to an evolution of said channel, (i.e., the broadest reasonable interpretation includes a step of evaluation and judgement and could be performed mentally or with pen and paper like marking nodes and edges with numerical values, which is either a mental process of evaluation/judgement (MPEP 2106)).
…establish a match between an evolution of the channel and the adaptation of the parameters of the first network and being used to determine parameters of said first network to be frozen and parameters of said first network to be adapted following an evolution of the channel; (i.e., the broadest reasonable interpretation includes a step of observation, evaluation, and judgement and could be performed mentally or with pen and paper like marking parameters that benefitted the model performance and marking parameters that did not benefit the model, which is either a mental process of observation/evaluation/judgement (MPEP 2106)).
and adapting the identified parameters of said first neural network by using the information on said evolution provided by said second neural network. (i.e., the broadest reasonable interpretation includes a step of observation, evaluation, and judgement and could be performed mentally or with pen and paper like updating parameters to improve model performance, which is either a mental process of observation/evaluation/judgement (MPEP 2106)).
If the claim limitations, under their broadest reasonable interpretation, covers activities classified under Mental processes: concepts performed in the human mind (including observation, evaluation, judgement, or opinion) (see MPEP 2106.04(a)(2), subsection (III)) or Mathematical concepts: mathematical relationships, mathematical formulas or equations, or mathematical calculations (see MPEP 2106.04(a)(2), subsection (I)). Accordingly, the claim recites an abstract idea.
In Step 2A, Prong 2 of the 101 analysis, set forth in MPEP 2106, the examiner has determined that the following additional elements do not integrate this judicial exception into a practical application:
used in a communication network to implement a processing by an equipment of an input signal received after transmission by a communication channel of a signal emitted by a terminal and to obtain an output signal, (i.e., the broadest reasonable interpretation of receiving a data instance is mere data gathering, which is an insignificant extra solution activity (MPEP 2106.05(g))).
the parameters of the first neural network depending on propagation characteristics of the communication channel, said first neural network has been trained for initial values of propagation characteristics of the communication channel, (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
sending by the first neural network information on said evolution to the second neural network, (i.e., the broadest reasonable interpretation of receiving a data instance is mere data gathering, which is an insignificant extra solution activity (MPEP 2106.05(g))).
receiving information, provided by said second neural network, that identifies parameters of said first network to be frozen and parameters of the first network to be adapted following the detected evolution; (i.e., the broadest reasonable interpretation of receiving a data instance is mere data gathering, which is an insignificant extra solution activity (MPEP 2106.05(g))).
said second neural network having been trained to… (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
Since the claim does not contain any other additional elements, that amount to integration into a practical application, the claim is directed to an abstract idea.
In Step 2B of the 101 analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception:
Regarding limitations (IV, VI, and VII), under the broadest reasonable interpretation, recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity when considering evidence in view of Berkheimer v. HP, Inc., 881 F.3d 1360, 1368, 125 USPQ2d 1649, 1654 (Fed. Cir. 2018), see USPTO Berkheimer Memorandum (April 2018)).
Examiner uses Berkheimer: Option 2, a citation to one or more of the court decisions discussed in MPEP 2106.05(d)(II) as noting well-understood, routine, and conventional nature of the additional elements:
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
Further, limitation (V), under the broadest reasonable interpretation, merely recite steps that apply a generic neural network on a communications network, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Similarly, limitation (VIII), under the broadest reasonable interpretation, merely recite steps that apply a generic neural network as a tool to perform a judicial exception, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 2, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 2 recites wherein the information on said evolution includes: values of propagation characteristics of the channel before the evolution, for which said first neural network is optimized; and estimated values of these propagation characteristics following the evolution. Under the broadest reasonable interpretation, the limitations recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity (MPEP 2106.05(g)). Therefore, claim 2 does not solve the deficiencies of claim 1.
Regarding claim 3, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 3 recites wherein the information on said evolution includes: values of propagation characteristics of the channel before the evolution, for which said first neural network is optimized; and estimated values of these propagation characteristics following the evolution. Under the broadest reasonable interpretation, the limitations recite adapting parameters of a model given historical data and a prediction which is a step of observation, evaluation, and judgement which can be performed mentally or with pen and paper. The steps of observation, evaluation, and judgement are mental processes. Therefore, claim 3 does not solve the deficiencies of claim 1.
Regarding claim 4, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 4 recites wherein the information on said evolution includes a complex difference between values of characteristics of propagation on said channel before the evolution and estimated values of these characteristics after the evolution. Under the broadest reasonable interpretation, the limitations recite performing a complex difference between values which is interpreted as using a mathematical calculation. A mathematical calculation is interpreted as a mathematical concept. Therefore, claim 4 does not solve the deficiencies of claim 1.
Regarding claim 5, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 5 recites wherein the detection of a degradation in the quality of the processing due to an evolution of the channel includes: a comparison between a threshold and a value of a variation of a characteristic of propagation of the received input signal; and/or a comparison, for a given input signal, between an output signal obtained by the processing and a reference signal. Under the broadest reasonable interpretation, the limitations recite comparing two values to determine drift which is a step of observation, evaluation, and judgement which can be performed mentally or with pen and paper. The steps of observation, evaluation, and judgement are mental processes. Therefore, claim 5 does not solve the deficiencies of claim 1.
Regarding claim 6, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 6 recites before said adaptation, a modification of the information provided by said second neural network and used for said adaptation. Under the broadest reasonable interpretation, the limitations recite changing values before sending them which is a step of observation, evaluation, and judgement which can be performed mentally or with pen and paper. The steps of observation, evaluation, and judgement are mental processes. Therefore, claim 6 does not solve the deficiencies of claim 1.
Regarding claim 7, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A method for determining parameters of a first neural network to be frozen or adapted. The claim recites a method. A method is one of the four statutory categories of invention.
In Step 2A, Prong 1 of the 101 analysis set forth in MPEP 2106, the examiner has determined that the following limitations recite a process that, under broadest reasonable interpretation, covers a mental process or mathematical concept but for the recitation of generic computer components:
… establish a match between an evolution of the channel and the adaptation of the parameters of the first network; (i.e., the broadest reasonable interpretation includes a step of observation, evaluation, and judgement and could be performed mentally or with pen and paper like marking parameters that benefitted the model performance and marking parameters that did not benefit the model, which is either a mental process of observation/evaluation/judgement (MPEP 2106)).
determining…from the received information, the parameters of said first network to be frozen and the parameters of said first network to be adapted following said evolution; (i.e., the broadest reasonable interpretation includes a step of observation, evaluation, and judgement and could be performed mentally or with pen and paper like marking parameters that benefitted the model performance and marking parameters that did not benefit the model, which is either a mental process of observation/evaluation/judgement (MPEP 2106)).
If the claim limitations, under their broadest reasonable interpretation, covers activities classified under Mental processes: concepts performed in the human mind (including observation, evaluation, judgement, or opinion) (see MPEP 2106.04(a)(2), subsection (III)) or Mathematical concepts: mathematical relationships, mathematical formulas or equations, or mathematical calculations (see MPEP 2106.04(a)(2), subsection (I)). Accordingly, the claim recites an abstract idea.
In Step 2A, Prong 2 of the 101 analysis, set forth in MPEP 2106, the examiner has determined that the following additional elements do not integrate this judicial exception into a practical application:
said first neural network being used in a communication network to implement a processing by an equipment of an input signal received after transmission by a communication channel of a signal emitted by a terminal and to obtain an output signal, (i.e., the broadest reasonable interpretation of receiving a data instance is mere data gathering, which is an insignificant extra solution activity (MPEP 2106.05(g))).
the parameters of the first neural network depending on propagation characteristics of the communication channel, said first neural network having been trained for initial values of propagation characteristics of the communication channel, (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
said method comprising: learning a second neural network to… (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
receiving by the second neural network information on an evolution of said channel resulting in a degradation in the quality of the processing; (i.e., the broadest reasonable interpretation of receiving a data instance is mere data gathering, which is an insignificant extra solution activity (MPEP 2106.05(g))).
…by said second neural network,… (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
and sending to the first neural network information that identifies the parameters to be frozen and the parameters to be adapted. (i.e., the broadest reasonable interpretation of receiving a data instance is mere data gathering, which is an insignificant extra solution activity (MPEP 2106.05(g))).
Since the claim does not contain any other additional elements, that amount to integration into a practical application, the claim is directed to an abstract idea.
In Step 2B of the 101 analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception:
Regarding limitations (III, VI, and VIII), under the broadest reasonable interpretation, recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity when considering evidence in view of Berkheimer v. HP, Inc., 881 F.3d 1360, 1368, 125 USPQ2d 1649, 1654 (Fed. Cir. 2018), see USPTO Berkheimer Memorandum (April 2018)).
Examiner uses Berkheimer: Option 2, a citation to one or more of the court decisions discussed in MPEP 2106.05(d)(II) as noting well-understood, routine, and conventional nature of the additional elements:
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
Further, limitation (IV), under the broadest reasonable interpretation, merely recite steps that apply a generic neural network on a communications network, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Similarly, limitation (V and VII), under the broadest reasonable interpretation, merely recite steps that apply a generic neural network as a tool to perform a judicial exception, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 8, it is dependent upon claim 7 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 8 recites wherein said information identifying the parameters to be frozen and the parameters to be adapted includes at least a rate of learning of a parameter of said first neural network and/or at least one weight value associated with a said learning rate. Under the broadest reasonable interpretation, the limitations recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity (MPEP 2106.05(g)). Therefore, claim 8 does not solve the deficiencies of claim 7.
Regarding claim 9, it is dependent upon claim 7 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 9 recites further comprising providing to the first neural network initial values of the parameters of the first network to be adapted. Under the broadest reasonable interpretation, the limitations recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity (MPEP 2106.05(g)). Therefore, claim 9 does not solve the deficiencies of claim 7.
Regarding claim 10, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 10 recites wherein said second neural network is trained by: values of the characteristics of said channel before and after an evolution; and parameters of said first neural network before and after the evolution and/or output signals of the processing associated with said values of the characteristics of the channel. Under the broadest reasonable interpretation, the limitations merely recite steps that apply historical training data to train a generic neural network, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Therefore, claim 10 does not solve the deficiencies of claim 1.
Regarding claim 11, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 11 recites wherein said second neural network is trained by: values of the characteristics of said channel before and after an evolution, the first neural network being trained for the characteristics of the channel before the evolution;. Under the broadest reasonable interpretation, the limitations merely recite steps that apply historical training data to train a generic neural network, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Claim 11 also recites sending to the first neural network of information that identifies the parameters, randomly determined, to be frozen and adapted; receipt of information on the quality of the processing of the first neural network after adaptation of its parameters according to the provided information;. Under the broadest reasonable interpretation, the limitations recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity (MPEP 2106.05(g)). Claim 11 also recites evaluation of influence of the information provided to determine new information to be provided. Under the broadest reasonable interpretation, the limitations recite determining if model parameters are good or not which is a step of observation, evaluation, and judgement which can be performed mentally or with pen and paper. The steps of observation, evaluation, and judgement are mental processes. Therefore, claim 11 does not solve the deficiencies of claim 1.
Regarding claim 12, it is dependent upon claim 10 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 12 recites wherein the second neural network receives, in association with the information on the evolution of the channel, at least one parameter of the first neural network among a weight and a bias, and/or a value of a loss function and/or at least one component of a gradient of the loss function. Under the broadest reasonable interpretation, the limitations recite steps of mere data gathering, which has been recognized by the courts as being well-understood, routine, and conventional functions. Specifically, the courts have recognized computer functions directed to mere data gathering as well-understood, routine, and conventional functions when they are claimed in a merely generic manner or as insignificant extra-solution activity (MPEP 2106.05(g)). Therefore, claim 12 does not solve the deficiencies of claim 10.
Regarding claim 13, it is dependent upon claim 12 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 13 recites wherein the second neural network uses a gradient back-propagation technique for its learning. Under the broadest reasonable interpretation, the limitations merely recite steps that apply generic training using back-propagation, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Therefore, claim 13 does not solve the deficiencies of claim 12.
Regarding claim 14, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 14 recites wherein the first neural network is trained by exploitation of received signals corresponding to a learning sequence emitted by a terminal, a first update of the parameters of the first neural network being performed during this training. Under the broadest reasonable interpretation, the limitations merely recite steps that apply generic training of a neural network, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Therefore, claim 14 does not solve the deficiencies of claim 1.
Regarding claim 15, it is dependent upon claim 14 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 15 recites wherein the parameters of the first neural network are updated iteratively by back-propagation of a gradient by minimizing a cost function based on a quality of the reconstruction at the end of the processing by the first neural network of the learning sequence. Under the broadest reasonable interpretation, the limitations merely recite steps that apply generic training using back-propagation, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Therefore, claim 15 does not solve the deficiencies of claim 14.
Regarding claim 16, it is dependent upon claim 1 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 16 recites wherein the processing is a function taken from among: an equalization function, and a signal processing function performing at least one time or frequency drift to maintain a time and/or frequency synchronization between an emitter and a receiver. Under the broadest reasonable interpretation, the limitations merely recite steps that apply generic signal processing functions, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Therefore, claim 16 does not solve the deficiencies of claim 1.
Regarding claim 17, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A non-transitory computer readable medium. The claim recites a computer readable medium which is interpreted as an article of manufacture. An article of manufacture is one of the four statutory categories of invention. For the Step 2A/2B analyses, since claim 17 is similar to claim 1 it is rejected under the same rationales as claim 1.
The additional limitation below fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception.
A non-transitory computer readable medium having stored thereon instructions which, when executed by a processor, cause the processor to (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 18, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A non-transitory computer readable medium. The claim recites a computer readable medium which is interpreted as an article of manufacture. An article of manufacture is one of the four statutory categories of invention. For the Step 2A/2B analyses, since claim 18 is similar to claim 7 it is rejected under the same rationales as claim 7.
The additional limitation below fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception.
A non-transitory computer readable medium having stored thereon instructions which, when executed by a processor, cause the processor to (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 19, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A device configured to. The claim recites a device which is interpreted as a machine. A machine is one of the four statutory categories of invention. For the Step 2A/2B analyses, since claim 19 is similar to claim 1 it is rejected under the same rationales as claim 1.
The additional limitation below fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception.
A device configured to… (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 20, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A system for adapting the parameters of a first neural network, said system including: at least one device configured to. The claim recites a system which is interpreted as a machine. A machine is one of the four statutory categories of invention. For the Step 2A/2B analyses, since claim 20 is similar to claim 1 it is rejected under the same rationales as claim 1.
The additional limitation below fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception.
A system for adapting the parameters of a first neural network, said system including: at least one device…and a second device (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 21, it is dependent upon claim 20 and fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. For example, claim 21 recites wherein said at least one first device is a base station or a terminal and said second device is a server of a core of said communication network. Under the broadest reasonable interpretation, the limitations merely recite steps that apply generic computer components, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)). Therefore, claim 21 does not solve the deficiencies of claim 1.
Regarding claim 22, in step 1 of the 101 analysis set forth in MPEP 2106, the claim recites A device configured to determine parameters of a first neural network to be frozen or adapted. The claim recites a device which is interpreted as a machine. A machine is one of the four statutory categories of invention. For the Step 2A/2B analyses, since claim 22 is similar to claim 7 it is rejected under the same rationales as claim 7.
The additional limitation below fails to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception.
A device configured to… (i.e., the generic computer components recited in this limitation merely add the words “apply it”, or an equivalent, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f))).
Considering additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
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.
Claims 1-15 and 17-22 are rejected under 35 U.S.C. 103 as being unpatentable over Vitthaladevuni, et al., US Pre-Grant Publication US20230188302A1 (“Vitthaladevuni”) in view of Mao, et al., Non-Patent Literature “RoemNet: Robust Meta Learning Based Channel Estimation in OFDM Systems” (“Mao”) and further in view of Tan, et al., US Pre-Grant Publication US20210232909A1 (“Tan”).
Regarding claim 1, Vitthaladevuni discloses:
A method for adapting parameters of a first neural network used in a communication network to implement a processing by an equipment of an input signal received after transmission by a communication channel of a signal emitted by a terminal and to obtain an output signal, (Vitthaladevuni, ⁋37, “As described herein, a wireless communications system may support efficient techniques that may allow a UE [used in a communication network to implement a processing by an equipment] to report CSI to a base station at an appropriate level of accuracy [of an input signal received after transmission by a communication channel of a signal emitted by a terminal and to obtain an output signal,]. In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) [A method for adapting parameters of a first neural network] based on a level of accuracy indicated by the base station.”).
the parameters of the first neural network depending on propagation characteristics of the communication channel, said first neural network has been trained for initial values of propagation characteristics of the communication channel, (Vitthaladevuni, ⁋37, “the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy with which the CSI is to be reported by the UE [the parameters of the first neural network depending on propagation characteristics of the communication channel,], and the UE may train the neural network pair using the loss metric or function [said first neural network has been trained for initial values of propagation characteristics of the communication channel,].”).
said method comprising: following a detection of a deterioration in the quality of the processing due to an evolution of said channel, sending by the first neural network information on said evolution… (Vitthaladevuni, ⁋37, “a wireless communications system may support efficient techniques that may allow a UE to report CSI [sending by the first neural network information on said evolution…] to a base station at an appropriate level of accuracy. In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy [said method comprising: following a detection of a deterioration in the quality of the processing due to an evolution of said channel,] with which the CSI is to be reported by the UE, and the UE may train the neural network pair using the loss metric or function.”, and Vitthaladevuni, ⁋4, “and the base station may use the CSI to improve the quality of downlink transmissions [detection of a deterioration in the quality] to the UE. For example, the CSI may include a channel quality indicator (CQI), and the base station may use the CQI to identify appropriate parameters (e.g., a modulation and coding scheme (MCS)) for transmitting downlink data to the UE.”).
receiving information,…that identifies…parameters of the first network to be adapted following the detected evolution; (Vitthaladevuni, ⁋37, “the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station [receiving information,…that identifies…parameters of the first network to be adapted following the detected evolution;].”).
and adapting the identified parameters of said first neural network by using the information on said evolution…. (Vitthaladevuni, ⁋37, “the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy with which the CSI is to be reported by the UE, and the UE may train the neural network pair using the loss metric or function [and adapting the identified parameters of said first neural network by using the information on said evolution….].”).
While Vitthaladevuni teaches a system that adapts a first channel neural network model using training parameters provided by a base station, Vitthaladevuni does not explicitly teach:
…to the second neural network, said second neural network having been trained to establish a match between an evolution of the channel and the adaptation of the parameters of the first network and being used to determine parameters of said first network to be frozen and parameters of said first network to be adapted following an evolution of the channel;
receiving information, provided by said second neural network, that identifies parameters of said first network to be frozen and parameters of the first network to be adapted following the detected evolution;
and adapting the identified parameters of said first neural network by using the information on said evolution provided by said second neural network.
Mao teaches:
…to the second neural network, said second neural network having been trained to establish a match between an evolution of the channel and the adaptation of the parameters of the first network and being used to determine…parameters of said first network to be adapted following an evolution of the channel; (Mao, pg. 3 col. 1, “In RoemNet […to the second neural network,], the training data of typical channels is acquired as multiple tasks to learn the general characteristics. When facing an unknown channel, RoemNet can use the K pilots to fine-tune the network [first network] through K-shot SGD [said second neural network having been trained to establish a match between an evolution of the channel and the adaptation of the parameters of the first network]. The core of RoemNet is to train the initialization parameters, and update them through several gradient steps with the help of a small amount of data from new tasks, so that the adaptation happens in the right space for fast learning on new channels [and being used to determine…parameters of said first network to be adapted following an evolution of the channel;].”).
receiving information, provided by said second neural network, that identifies…parameters of the first network to be adapted following the detected evolution; (Mao, pg. 4 col. 1 and Algorithm 2, “The update of parameters is not done like common NN with standard supervised learning method, but with all the rules learned from the priori channels, just like priori guidance. This process is called meta-learner. As shown in Fig. 3, during training, the standard supervised learning based NN updates the parameters every time in which there is a gradient descent. Differently, RoemNet firstly calculate all the gradients of loss function using Eq. (6). Next with Eq. (7), RoemNet can find a group of appropriate initial parameters that has the best generalization ability for different channel tasks in the parameter space [receiving information, provided by said second neural network, that identifies…parameters of the first network to be adapted following the detected evolution;].”).
and adapting the identified parameters of said first neural network by using the information on said evolution provided by said second neural network. (Mao, pg. 2 col. 1, “Then the meta-learner will be used to guide the adjustment of a new network for its deployment in unknown channels, where few pilot symbols are available for fine tuning. This process is called meta-update. The prior guidance together with meta-update make RoemNet possible to be robust enough to achieve rapid convergence with few pilots [and adapting the identified parameters of said first neural network by using the information on said evolution provided by said second neural network.].”).
Vitthaladevuni and Mao are both in the same field of endeavor (i.e. neural networks). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Vitthaladevuni and Mao to teach the above limitation(s). The motivation for doing so is that using meta learning improves the robustness of a base model to changing environments (cf. Mao, pg. 2 col. 1, “With meta learning (or learning-to-learn) [15] [16], RoemNet is able to acquire many skills and adapt to many environments. Therefore, it can make a quick solution to new learning tasks using only a small number of training samples. Unlike traditional learning methods that learn an end to-end mapping with trained weights, meta-net is trained to learn meta knowledge, i.e., various basic parameters during neural network training, such as initialization of parameters [17], the choice of optimizers [18], even the structure of models [19], etc.”).
While Vitthaladevuni in view of Mao teaches a system for adapting a first channel neural network using a second neural network, the combination does not explicitly teach:
…determine parameters of said first network to be frozen…
Tan teaches …determine parameters of said first network to be frozen… (Tan, ⁋6, “a selection component, and a freeze-out component. The assessment component identifies units of a neural network. The selection component selects a subset of units of the neural network. The freeze-out component freezes the selected subset of units of the neural network so that weights of output connections from the frozen subset of units will not be updated for a training run […determine parameters of said first network to be frozen…].”).
Vitthaladevuni, in view of Mao, and Tan are both in the same field of endeavor (i.e. neural networks). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Vitthaladevuni, in view of Mao, and Tan to teach the above limitation(s). The motivation for doing so is that freezing select parameters of a model during training increases robustness by reducing overfitting (cf. Tan, ⁋2, “Overfitting occurs during the training process when the neural network model learns the training data too well, resulting in lower performance when the model is presented with new, unseen data. Overfitting can be detected by applying validation metrics such as accuracy to the new, unseen data (test data) and the training data.”).
Regarding claim 2, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Vitthaladevuni further teaches further comprising sending, in association with the information on said evolution, at least one parameter of said first neural network among a weight and a bias, and/or a value of a loss function and/or at least one component of a gradient of the loss function. (Vitthaladevuni, ⁋70, “Further, because the UE 115 may train both the neural network at an encoder and a neural network at a decoder (e.g., the encoder and decoder may be at the UE 115, and the decoder may also be at a base station 105), the UE 115 may send coefficients of the decoder neural network to a base station 105 [further comprising sending, in association with the information on said evolution, at least one parameter of said first neural network among a weight and a bias,].”).
Regarding claim 3, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Vitthaladevuni further teaches wherein the information on said evolution includes: values of propagation characteristics of the channel before the evolution, for which said first neural network is optimized; and estimated values of these propagation characteristics following the evolution. (Vitthaladevuni, ⁋61, “In wireless communications system 100, a UE 115 may be configured to perform channel measurements on downlink signals received from a base station 105 and report the channel measurements to the base station 105. The UE 115 may report the channel measurements as CSI feedback [wherein the information on said evolution includes: values of propagation characteristics of the channel before the evolution, for which said first neural network is optimized;].”, and Vitthaladevuni, ⁋37, “For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy with which the CSI is to be reported by the UE, and the UE may train the neural network pair using the loss metric or function. As such, the UE may report the CSI at the level of accuracy indicated by the base station [and estimated values of these propagation characteristics following the evolution.].”).
Regarding claim 4, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Vitthaladevuni further teaches wherein the information on said evolution includes a complex difference between values of characteristics of propagation on said channel before the evolution and estimated values of these characteristics after the evolution. (Vitthaladevuni, ⁋77, “For example, a high level of accuracy may indicate a small or no difference between the raw CSI entered (or the actual channel condition measured) [wherein the information on said evolution includes a complex difference between values of characteristics of propagation on said channel before the evolution] as input to the encoder and the CSI produced by the decoder [and estimated values of these characteristics after the evolution.]”).
Regarding claim 5, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Vitthaladevuni further teaches wherein the detection of a degradation in the quality of the processing due to an evolution of the channel includes: a comparison between a threshold and a value of a variation of a characteristic of propagation of the received input signal; and/or a comparison, for a given input signal, between an output signal obtained by the processing and a reference signal. (Vitthaladevuni, ⁋37, “a wireless communications system may support efficient techniques that may allow a UE to report CSI to a base station at an appropriate level of accuracy. In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy [wherein the detection of a degradation in the quality of the processing due to an evolution of the channel includes:] with which the CSI is to be reported by the UE, and the UE may train the neural network pair using the loss metric or function.”, and Vitthaladevuni, ⁋77, “For example, a high level of accuracy may indicate a small or no difference between the raw CSI entered (or the actual channel condition measured) [a comparison between a threshold and a value of a variation of a characteristic of propagation of the received input signal;] as input to the encoder and the CSI produced by the decoder”).
Regarding claim 6, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Mao further teaches further comprising, before said adaptation, a modification of the information provided by said second neural network and used for said adaptation. (Mao, pg. 4 col. 1-2, “When faced with blocks for testing, there is an update for meta-learner to make RoemNet adapted to time varying channels [further comprising, before said adaptation,]. Given that there are only K2 known pilot signals in one block, we set K2 pilots as one group. Different from training period, since OFDM communication system calls for real-time transmission, we use SGD for meta-update, which requires relatively few computing [a modification of the information provided by said second neural network and used for said adaptation.]”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 1.
Regarding claim 7, Vitthaladevuni discloses:
A method for determining parameters of a first neural network to be frozen or adapted, said first neural network being used in a communication network to implement a processing by an equipment of an input signal received after transmission by a communication channel of a signal emitted by a terminal and to obtain an output signal, (Vitthaladevuni, ⁋37, “As described herein, a wireless communications system may support efficient techniques that may allow a UE [used in a communication network to implement a processing by an equipment] to report CSI to a base station at an appropriate level of accuracy [of an input signal received after transmission by a communication channel of a signal emitted by a terminal and to obtain an output signal,]. In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) [A method for determining parameters of a first neural network to be frozen or adapted,] based on a level of accuracy indicated by the base station.”).
the parameters of the first neural network depending on propagation characteristics of the communication channel, said first neural network having been trained for initial values of propagation characteristics of the communication channel, (Vitthaladevuni, ⁋37, “the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy with which the CSI is to be reported by the UE [the parameters of the first neural network depending on propagation characteristics of the communication channel,], and the UE may train the neural network pair using the loss metric or function [said first neural network having been trained for initial values of propagation characteristics of the communication channel,].”).
receiving…information on an evolution of said channel resulting in a degradation in the quality of the processing; (Vitthaladevuni, ⁋37, “a wireless communications system may support efficient techniques that may allow a UE to report CSI to a base station [receiving…information on an evolution of said channel] at an appropriate level of accuracy. In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy [resulting in a degradation in the quality of the processing;] with which the CSI is to be reported by the UE, and the UE may train the neural network pair using the loss metric or function.”, and Vitthaladevuni, ⁋4, “and the base station may use the CSI to improve the quality of downlink transmissions [resulting in a degradation in the quality of the processing;] to the UE. For example, the CSI may include a channel quality indicator (CQI), and the base station may use the CQI to identify appropriate parameters (e.g., a modulation and coding scheme (MCS)) for transmitting downlink data to the UE.”).
determining…, from the received information,…the parameters of said first network to be adapted following said evolution; (Vitthaladevuni, ⁋37, “the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station [determining…, from the received information,…the parameters of said first network to be adapted following said evolution;].”).
and sending to the first neural network information that identifies…the parameters to be adapted. (Vitthaladevuni, ⁋37, “the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station. For instance, the base station may indicate a loss metric or function corresponding to the level of accuracy with which the CSI is to be reported by the UE, and the UE may train the neural network pair using the loss metric or function [and sending to the first neural network information that identifies…the parameters to be adapted.].”).
While Vitthaladevuni teaches a system that adapts a first channel neural network model using training parameters provided by a base station, Vitthaladevuni does not explicitly teach:
according to a method comprising: learning a second neural network to establish a match between an evolution of the channel and the adaptation of the parameters of the first network;
receiving by the second neural network information on an evolution of said channel resulting in a degradation in the quality of the processing;
determining by said second neural network, from the received information, the parameters of said first network to be frozen and the parameters of said first network to be adapted following said evolution;
and sending to the first neural network information that identifies the parameters to be frozen and the parameters to be adapted.
Mao teaches:
said method comprising: learning a second neural network to establish a match between an evolution of the channel and the adaptation of the parameters of the first network; (Mao, pg. 3 col. 1, “In RoemNet [second neural network], the training data of typical channels is acquired as multiple tasks to learn the general characteristics. When facing an unknown channel, RoemNet can use the K pilots to fine-tune the network [first network] through K-shot SGD [said method comprising: learning a second neural network to establish a match between an evolution of the channel and the adaptation of the parameters of the first network;]. The core of RoemNet is to train the initialization parameters, and update them through several gradient steps with the help of a small amount of data from new tasks, so that the adaptation happens in the right space for fast learning on new channels.”).
receiving by the second neural network information on an evolution of said channel resulting in a degradation in the quality of the processing; (Mao, pg. 2 col. 1, “Then the meta-learner will be used to guide the adjustment of a new network for its deployment in unknown channels, where few pilot symbols are available for fine tuning. This process is called meta-update. The prior guidance together with meta-update make RoemNet possible to be robust enough to achieve rapid convergence with few pilots [receiving by the second neural network information on an evolution of said channel resulting in a degradation in the quality of the processing;].”).
determining by said second neural network, from the received information,…and the parameters of said first network to be adapted following said evolution; (Mao, pg. 4 col. 1 and Algorithm 2, “The update of parameters is not done like common NN with standard supervised learning method, but with all the rules learned from the priori channels, just like priori guidance. This process is called meta-learner. As shown in Fig. 3, during training, the standard supervised learning based NN updates the parameters every time in which there is a gradient descent. Differently, RoemNet firstly calculate all the gradients of loss function using Eq. (6). Next with Eq. (7), RoemNet can find a group of appropriate initial parameters that has the best generalization ability for different channel tasks in the parameter space [determining by said second neural network, from the received information,…and the parameters of said first network to be adapted following said evolution;].”).
Vitthaladevuni and Mao are both in the same field of endeavor (i.e. neural networks). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Vitthaladevuni and Mao to teach the above limitation(s). The motivation for doing so is that using meta learning improves the robustness of a base model to changing environments (cf. Mao, pg. 2 col. 1, “With meta learning (or learning-to-learn) [15] [16], RoemNet is able to acquire many skills and adapt to many environments. Therefore, it can make a quick solution to new learning tasks using only a small number of training samples. Unlike traditional learning methods that learn an end to-end mapping with trained weights, meta-net is trained to learn meta knowledge, i.e., various basic parameters during neural network training, such as initialization of parameters [17], the choice of optimizers [18], even the structure of models [19], etc.”).
While Vitthaladevuni in view of Mao teaches a system for adapting a first channel neural network using a second neural network, the combination does not explicitly teach:
…the parameters of said first network to be frozen…
Tan teaches …the parameters of said first network to be frozen… (Tan, ⁋6, “a selection component, and a freeze-out component. The assessment component identifies units of a neural network. The selection component selects a subset of units of the neural network. The freeze-out component freezes the selected subset of units of the neural network so that weights of output connections from the frozen subset of units will not be updated for a training run […the parameters of said first network to be frozen…].”).
Vitthaladevuni, in view of Mao, and Tan are both in the same field of endeavor (i.e. neural networks). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Vitthaladevuni, in view of Mao, and Tan to teach the above limitation(s). The motivation for doing so is that freezing select parameters of a model during training increases robustness by reducing overfitting (cf. Tan, ⁋2, “Overfitting occurs during the training process when the neural network model learns the training data too well, resulting in lower performance when the model is presented with new, unseen data. Overfitting can be detected by applying validation metrics such as accuracy to the new, unseen data (test data) and the training data.”).
Regarding claim 8, Vitthaladevuni in view of Mao and Tan teaches the method of claim 7. Tan also teaches the parameters to be frozen as seen in claim 1. Mao further teaches wherein said information identifying…and the parameters to be adapted includes at least a rate of learning of a parameter of said first neural network and/or at least one weight value associated with a said learning rate. (Mao, pg. 4 col. 1-2, “When faced with blocks for testing, there is an update for meta-learner to make RoemNet adapted to time varying channels [wherein said information identifying…the parameters to be adapted]. Given that there are only K2 known pilot signals in one block, we set K2 pilots as one group. Different from training period, since OFDM communication system calls for real-time transmission, we use SGD for meta-update, which requires relatively few computing; using SGD to update the base model’s weights is interpreted as a weight value associated with a learning rate (i.e. includes…and/or at least one weight value associated with a said learning rate.)”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 7.
Regarding claim 9, Vitthaladevuni in view of Mao and Tan teaches the method of claim 7. Mao further teaches further comprising providing to the first neural network initial values of the parameters of the first network to be adapted. (Mao, pg. 2 col. 1, “In RoemNet, we consider multiple channels as multiple tasks, and we train a meta-learner that can learn the general characteristics of these tasks to get a good initialization for the network [further comprising providing to the first neural network initial values of the parameters of the first network to be adapted.].”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 7.
Regarding claim 10, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Mao further teaches wherein said second neural network is trained by: values of the characteristics of said channel before and after an evolution; and parameters of said first neural network before and after the evolution and/or output signals of the processing associated with said values of the characteristics of the channel. (Mao, pg. 3 col. 1, “In RoemNet, the training data of typical channels is acquired as multiple tasks to learn the general characteristics. When facing an unknown channel, RoemNet can use the K pilots to fine-tune the network through K-shot SGD [wherein said second neural network is trained by: values of the characteristics of said channel before and after an evolution;]. The core of RoemNet is to train the initialization parameters, and update them through several gradient steps with the help of a small amount of data from new tasks, so that the adaptation happens in the right space for fast learning on new channels [and parameters of said first neural network before and after the evolution….].”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 1.
Regarding claim 11, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Tan also teaches parameters to be frozen as seen in claim 1. Mao further teaches:
wherein said second neural network is trained by: values of the characteristics of said channel before and after an evolution, the first neural network being trained for the characteristics of the channel before the evolution; (Mao, pg. 3 col. 1, “In RoemNet, the training data of typical channels is acquired as multiple tasks to learn the general characteristics. When facing an unknown channel, RoemNet can use the K pilots to fine-tune the network through K-shot SGD [wherein said second neural network is trained by: values of the characteristics of said channel before and after an evolution,].”, and Mao, pg. 2 col. 1, “In RoemNet, we consider multiple channels as multiple tasks, and we train a meta-learner that can learn the general characteristics of these tasks to get a good initialization for the network [the first neural network being trained for the characteristics of the channel before the evolution;].”).
sending to the first neural network of information that identifies the parameters, randomly determined,…and adapted; (Mao, pg. 2 col. 1 and Algorithm 1, “Then the meta-learner will be used to guide the adjustment of a new network for its deployment in unknown channels, where few pilot symbols are available for fine tuning. This process is called meta-update. The prior guidance together with meta-update make RoemNet possible to be robust enough to achieve rapid convergence with few pilots [sending to the first neural network of information that identifies the parameters, randomly determined,…and adapted;].”).
receipt of information on the quality of the processing of the first neural network after adaptation of its parameters according to the provided information; (Mao, pg. 3 col. 1, “Secondly, frames are sent and received from an OFDM system over the time varying channel. Each block in one frame contains two parts, the pilots and the information. The received and transmitted pilot data of one block is collected and saved as the input and output of the meta net, respectively, for further adjustment [receipt of information on the quality of the processing of the first neural network after adaptation of its parameters according to the provided information;].”).
evaluation of influence of the information provided to determine new information to be provided. (Mao, pg. 3 col. 1, “Lastly, RoemNet uses the collected data to update the parameters and fine-tune the previous network by performing certain steps of Stochastic Gradient Descent (SGD) [evaluation of influence of the information provided to determine new information to be provided.].”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 1.
Regarding claim 12, Vitthaladevuni in view of Mao and Tan teaches the method of claim 10. Mao further teaches wherein the second neural network receives, in association with the information on the evolution of the channel, at least one parameter of the first neural network among a weight and a bias, and/or a value of a loss function and/or at least one component of a gradient of the loss function. (Mao, pg. 2 col. 1, “With meta learning (or learning-to-learn) [15] [16], RoemNet is able to acquire many skills and adapt to many environments. Therefore, it can make a quick solution to new learning tasks using only a small number of training samples. Unlike traditional learning methods that learn an end to-end mapping with trained weights, meta-net is trained to learn meta knowledge, i.e., various basic parameters during neural network training, such as initialization of parameters [17], the choice of optimizers [18], even the structure of models [19], etc [wherein the second neural network receives, in association with the information on the evolution of the channel, at least one parameter of the first neural network among a weight and a bias….].”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 1.
Regarding claim 13, Vitthaladevuni in view of Mao and Tan teaches the method of claim 12. Mao further teaches wherein the second neural network uses a gradient back-propagation technique for its learning. (Mao, pg. 3 col. 1, “Lastly, RoemNet uses the collected data to update the parameters and fine-tune the previous network by performing certain steps of Stochastic Gradient Descent (SGD) [wherein the second neural network uses a gradient back-propagation technique for its learning.].”).
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Mao with the teachings of Vitthaladevuni and Tan for the same reasons disclosed in claim 1.
Regarding claim 14, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Vitthaladevuni further teaches wherein the first neural network is trained by exploitation of received signals corresponding to a learning sequence emitted by a terminal, a first update of the parameters of the first neural network being performed during this training. (Vitthaladevuni, ⁋62, “a UE 115 may input channel realization information into the encoder 205, and the encoder 205 may encode the channel realization information using the first neural network to generate CSI feedback. The channel realization information may refer to the raw channel and may correspond to measurements performed on CSI reference signals (CSI-RSs) received on the channel [wherein the first neural network is trained by exploitation of received signals corresponding to a learning sequence emitted by a terminal,].”, and Vitthaladevuni, ⁋65, “The neural network training may be iterative, such that the UE 115 trains a neural network based on a current version of the neural network and measurements attained since the current version of the neural network was implemented [a first update of the parameters of the first neural network being performed during this training.]”).
Regarding claim 15, Vitthaladevuni in view of Mao and Tan teaches the method of claim 14. Vitthaladevuni further teaches wherein the parameters of the first neural network are updated iteratively by back-propagation of a gradient by minimizing a cost function based on a quality of the reconstruction at the end of the processing by the first neural network of the learning sequence. (Vitthaladevuni, ⁋68, “The neural network pair may also implement a loss function, or cost function, based on the difference between an actual value and a predicted value [based on a quality of the reconstruction at the end of the processing by the first neural network of the learning sequence.]. For each layer of the neural network pair, the cost function may be used to adjust the weights for the next input based on a loss metric [wherein the parameters of the first neural network are updated iteratively…by minimizing a cost function].”, and Vitthaladevuni, ⁋63, “the UE 115 may implement deep learning (e.g., using a deep recurrent network), backpropagation [by back-propagation of a gradient]”).
Regarding claim 17, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. Vitthaladevuni teaches the additional limitation A non-transitory computer readable medium having stored thereon instructions which, when executed by a processor, cause the processor to (Vitthaladevuni, ⁋9, “A non-transitory computer-readable medium storing code for wireless communication at a UE is described. The code may include instructions executable by a processor to [A non-transitory computer readable medium having stored thereon instructions which, when executed by a processor, cause the processor to]”).
Regarding claim 18, Vitthaladevuni in view of Mao and Tan teaches the method of claim 7. Vitthaladevuni teaches the additional limitation A non-transitory computer-readable medium having stored thereon instructions which, when executed by a processor, cause the processor to (Vitthaladevuni, ⁋9, “A non-transitory computer-readable medium storing code for wireless communication at a UE is described. The code may include instructions executable by a processor to [A non-transitory computer-readable medium having stored thereon instructions which, when executed by a processor, cause the processor to]”).
Regarding claim 19, the claim is similar to claim 1 and rejected under the same rationales. Vitthaladevuni teaches the additional limitation A device configured to: adapt parameters of a first neural network (Vitthaladevuni, ⁋37, “In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station [A device configured to: adapt parameters of a first neural network].”).
Regarding claim 20, the claim is similar to claim 1 and rejected under the same rationales. Vitthaladevuni teaches the additional limitations:
A system for adapting the parameters of a first neural network, said system including: at least one device configured to adapt parameters of a first neural network (Vitthaladevuni, ⁋37, “In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station [A system for adapting the parameters of a first neural network, said system including: at least one device configured to adapt parameters of a first neural network].”).
and a second device according to claim 22, the second device having a higher computing capacity than that of said at least one first device. (Vitthaladevuni, ⁋93, “FIG. 7 shows a block diagram 700 of a communications manager 705 that supports configurable metrics for channel state compression and feedback in accordance with aspects of the present disclosure [and a second device according to claim 22,].”, and Vitthaladevuni, ⁋86, “In some examples, the communications manager 515, or its sub-components, may be combined with one or more other hardware components, including but not limited to an input/output (I/O) component, a transceiver, a network server; using a network server is interpreted as having more computing power than a base station (i.e. the second device having a higher computing capacity than that of said at least one first device.)”).
Regarding claim 21, Vitthaladevuni in view of Mao and Tan teaches the system of claim 20. Vitthaladevuni further teaches wherein said at least one first device is a base station or a terminal and said second device is a server of a core of said communication network. (Vitthaladevuni, ⁋37, “In particular, the UE may be configured to train a neural network pair (e.g., including a first neural network at an encoder and a second neural network at a decoder) based on a level of accuracy indicated by the base station [wherein said at least one first device is a base station or a terminal].”, and Vitthaladevuni, ⁋86, “In some examples, the communications manager 515, or its sub-components, may be combined with one or more other hardware components, including but not limited to an input/output (I/O) component, a transceiver, a network server [and said second device is a server of a core of said communication network.]”).
Regarding claim 22, the claim is similar to claim 7 and rejected under the same rationales. Vitthaladevuni teaches the additional limitation A device configured to determine parameters of a first neural network to be frozen or adapted, (Vitthaladevuni, ⁋93, “FIG. 7 shows a block diagram 700 of a communications manager 705 that supports configurable metrics for channel state compression and feedback in accordance with aspects of the present disclosure [A device configured to determine parameters of a first neural network to be frozen or adapted,].”, and Vitthaladevuni, ⁋86, “In some examples, the communications manager 515, or its sub-components, may be combined with one or more other hardware components, including but not limited to an input/output (I/O) component, a transceiver, a network server [A device]”).
Claim 16 is rejected under 35 U.S.C. 103 as being unpatentable over Vitthaladevuni, et al., US Pre-Grant Publication US20230188302A1 (“Vitthaladevuni”) in view of Mao, et al., Non-Patent Literature “RoemNet: Robust Meta Learning Based Channel Estimation in OFDM Systems” (“Mao”) and further in view of Tan, et al., US Pre-Grant Publication US20210232909A1 (“Tan”) and Li, et al., Non-Patent Literature “Convolutional recurrent neural network-based channel equalization: An experimental study” (“Li”).
Regarding claim 16, Vitthaladevuni in view of Mao and Tan teaches the method of claim 1. While the combination teaches a system for adapting a first channel neural network using a second neural network, the combination does not explicitly teach wherein the processing is a function taken from among: an equalization function, and a signal processing function performing at least one time or frequency drift to maintain a time and/or frequency synchronization between an emitter and a receiver.
Li teaches wherein the processing is a function taken from among: an equalization function, and a signal processing function performing at least one time or frequency drift to maintain a time and/or frequency synchronization between an emitter and a receiver. (Li, pg. 1 col. 2, “In this paper, we propose a CRNN-based channel equalizer [wherein the processing is a function taken from among: an equalization function,] which addresses the problem of temporal variations of data as well as nonlinear channel distortions. Our insight is leveraging the shift-invariant properties; shift-invariant is interpreted as synching the time drift between input and output signals (i.e. and a signal processing function performing at least one time or frequency drift to maintain a time and/or frequency synchronization between an emitter and a receiver.) of the convolutional neural network (CNN) [11] to learn matched filters analogous to the tap weights of conventional equalizer.”).
Vitthaladevuni, in view of Mao and Tan, and Li are both in the same field of endeavor (i.e. channel communications). It would have been obvious for a person having ordinary skill in the art before the effective filing date of the claimed invention to combine Vitthaladevuni, in view of Mao and Tan, and Li to teach the above limitation(s). The motivation for doing so is that using a channel equalizer and signal processing improves performance by reducing interference (cf. Li, pg. 1 col. 1, “The problem of equalization is to reconstruct the transmitted sequences and counteract the effects of ISI and noise based on the channel observations.”).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Zhang, et al., US20220147818A1 discloses a system that uses an auxiliary model to train a first neural network.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NICHOLAS S WU whose telephone number is (571)270-0939. The examiner can normally be reached Monday - Friday 8:00 am - 4:00 pm EST.
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, Michelle Bechtold can be reached at 571-431-0762. 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.
/N.S.W./Examiner, Art Unit 2148
/MICHELLE T BECHTOLD/Supervisory Patent Examiner, Art Unit 2148
Prosecution Insights