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 .
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicants’ submission filed on 10/27/25 has been entered.
DETAILED ACTION
The instant application having Application No. 17734510 has a total of 20 claims pending in the application, of which claims 5 and 17 have been cancelled.
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 1-11 and 13-18 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1 is a machine type claim. Claims 13 is a process type claim. Therefore, claims 1-11 and 13-18 are directed to either a process, machine, manufacture or composition of matter.
As per claim 1,
2A Prong 1:
“select a subset based on the reliability values for the user equipments and the reliability values for the training data set” The user mentally or with pencil and paper determines which data to use based upon available values.
2A Prong 2: This judicial exception is not integrated into a practical application.
Additional elements:
“an apparatus” “one processing core” “at least one memory” (mere instructions to apply the exception using a generic computer component);
“direct the subset of the group of user equipments to separately perform a machine learning training process in the user equipments in the subset” (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner’s note: Claims show no more than a generic machine learning algorithm across multiple generic devices, with no additional detail or limitations beyond a generic, off the shelf machine learning algorithm.
“obtain reliability values for each user equipment in a group of user equipments” , “for a respective user equipment in the group, send a request for a set of reliability values for a training data set stored in the respective user equipment, wherein the request comprises a respective metric”, “obtain, for each user equipment in the group, a reliability value, based at least in part on the requested metric, from the set of reliability values for the training data set stored in the user equipment” (Adding insignificant extra-solution activity to the judicial exception - see MPEP 2106.05(g)).
“wherein (i) each user equipment stores a distinct training data set and (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior” (Adding insignificant extra-solution activity to the judicial exception - see MPEP 2106.05(g)).
2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
“an apparatus” “one processing core” “at least one memory” (mere instructions to apply the exception using a generic computer component)
“direct the subset of the group of user equipments to separately perform a machine learning training process in the user equipments in the subset” (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner’s note: Claims show no more than a generic machine learning algorithm across multiple generic devices, with no additional detail or limitations beyond a generic, off the shelf machine learning algorithm.
“obtain reliability values for each user equipment in a group of user equipments” , “for a respective user equipment in the group, send a request for a set of reliability values for a training data set stored in the respective user equipment, wherein the request comprises a respective metric”, “obtain, for each user equipment in the group, a reliability value, based at least in part on the requested metric, from the set of reliability values for the training data set stored in the user equipment” (MPEP 2106.05(d)(II) indicate that merely “receiving and transmitting data” is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed obtaining step is well-understood, routine, conventional activity is supported under Berkheimer).
“wherein (i) each user equipment stores a distinct training data set and (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior” (MPEP 2106.05(d)(II) indicate that merely “storing and retrieving data in memory” is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed storing step is well-understood, routine, conventional activity is supported under Berkheimer).
As per claim 2, 4-5 and 9-10, these claims contain additional receiving steps, and are rejected for similar reasons to claim 1.
As per claim 3,
2A Prong 1:
“aggrege the results… to obtain an aggregated … result” The user mentally or with pencil and paper aggregates the data from the different sources.
The remaining machine learning and receiving steps of claim 3 are similar to those of claim 1, and rejected for similar reasons.
As per claim 6,
2A Prong 1:
“select the subset from among the group based on the compound reliability values of the user equipments of the group” The user mentally or with pencil and paper selects based upon the obtained values.
The remaining machine learning and receiving steps of claim 6 are similar to those of claim 1, and rejected for similar reasons.
As per claims 7-8, this claim contains similar storing steps to claim 1, and is rejected for similar reasons.
As per claim 11,
2A Prong 1:
“notify … comprised in the group but not comprised in the subset, that they have been excluded from the … process” The user mentally or with pencil and paper removes the unneeded portions and notes that they have been excluded.
The remaining machine learning and receiving steps of claim 11 are similar to those of claim 1, and rejected for similar reasons.
As per claim 13,
2A Prong 1:
“select a subset based on the reliability values for the user equipments and the reliability values for the training data sets” The user mentally or with pencil and paper determines which data to use based upon available values.
2A Prong 2: This judicial exception is not integrated into a practical application.
Additional elements:
“an apparatus” (mere instructions to apply the exception using a generic computer component);
“direct the subset of the group of user equipments to separately perform a machine learning training process in the user equipment in the subset” (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner’s note: Claims show no more than a generic machine learning algorithm across multiple generic devices, with no additional detail or limitations beyond a generic, off the shelf machine learning algorithm.
“obtaining … reliability values for each user equipment in a group of user equipments”, “For a respective user equipment in the group, sending a request to receive a set of reliability values for a training data set stored in the respective user equipment, wherein the request comprises a requested metric”, “obtaining, for each user equipment in the group, a reliability value based at least in part on the requested metric, from the set of reliability values for the training data set stored in the user equipment” (Adding insignificant extra-solution activity to the judicial exception - see MPEP 2106.05(g)).
“wherein (i) each user equipment stores a distinct training data set and (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior” (Adding insignificant extra-solution activity to the judicial exception - see MPEP 2106.05(g)).
2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
“an apparatus” (mere instructions to apply the exception using a generic computer component)
“direct the subset of the group of user equipments to separately perform a machine learning training process in the user equipment in the subset” (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f) – Examiner’s note: Claims show no more than a generic machine learning algorithm across multiple generic devices, with no additional detail or limitations beyond a generic, off the shelf machine learning algorithm.
“obtaining … reliability values for each user equipment in a group of user equipments”, “For a respective user equipment in the group, sending a request to receive a set of reliability values for a training data set stored in the respective user equipment, wherein the request comprises a requested metric”, “obtaining, for each user equipment in the group, a reliability value based at least in part on the requested metric, from the set of reliability values for the training data set stored in the user equipment” (MPEP 2106.05(d)(II) indicate that merely “receiving and transmitting data” is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed obtaining step is well-understood, routine, conventional activity is supported under Berkheimer).
“wherein (i) each user equipment stores a distinct training data set and (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior” (MPEP 2106.05(d)(II) indicate that merely “storing and retrieving data in memory” is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim). Thereby, a conclusion that the claimed storing step is well-understood, routine, conventional activity is supported under Berkheimer).
As per claim 14 and 16, these claims contain additional receiving steps, and are rejected for similar reasons to claim 13.
As per claim 15,
2A Prong 1:
“aggregating the results… to obtain an aggregate … result” The user mentally or with pencil and paper aggregates the data from the different sources.
The remaining machine learning and receiving steps of claim 3 are similar to those of claim 13, and rejected for similar reasons.
As per claim 18,
2A Prong 1:
“selecting the subset from among the group based on the compound reliability values of the user equipments of the group” The user mentally or with pencil and paper selects based upon the obtained values.
The remaining machine learning and receiving steps of claim 6 are similar to those of claim 13, and rejected for similar reasons.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-4, 6-12, 13-16, and 18-19 rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
As per claims 1 and 13, these claims call for “for a respective user equipment in the group, send a request for a set of reliability values for a training data set stored in the respective user equipment, wherein the request comprises a requested metric” and “Obtain, for each user equipment in the group, a reliability value, based at least in part on the requested metric, … wherein … (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior.” However, the specification does not support these limitations. The first part of this limitation requires a request for reliability values for a training data set, wherein the request comprises a requested metric. The second part requires that the reliability value indicate whether the respective user equipment is associated with anomalous behavior. Anomalous behavior is only discussed in two paragraphs in the specification, paragraph 0029 which denotes that the UE (User equipment) not be associated with reports of anomalous behavior, and to exclude them if they are found to be related to anomalous behavior. Paragraph 0049 mentions anomalous conditions, but does not elaborate on them being related to reliability values or the training data.
As can be seen, the specification at no time describes obtaining a reliability value for a UE based on a requested metric for training data that includes whether or not the UE has had anomalous behavior related to the training data. As the claim explicitly requires that the request be related to a set of reliability values for a training data set, and comprises a requested metric, the claim requires the anomalous behavior of the user equipment to be related to the training data and that metric, which is not supported by the specification. This causes these limitations to be new matter, and therefore rejected under U.S.C. 112(a).
As per claims 2-4, 6-11, 14-16 and 18, these claims are rejected as being dependent on a claim rejected under U.S.C. 112(a) for new matter.
As per claims 12 and 19, these claims call for “provide, responsive to a request comprising a requested metric … a reliability value for the set of training data … wherein the reliability value (i) is based on the requested metric and (ii) indicates whether the apparatus is associated with anomalous behavior.” However, the specification does not support these limitations. The first part of this limitation requires a request for reliability values for a training data set, wherein the request comprises a requested metric. The second part requires that the reliability value indicate whether the respective user equipment is associated with anomalous behavior. Anomalous behavior is only discussed in two paragraphs in the specification, paragraph 0029 which denotes that the UE (User equipment) not be associated with reports of anomalous behavior, and to exclude them if they are found to be related to anomalous behavior. Paragraph 0049 mentions anomalous conditions, but does not elaborate on them being related to reliability values or the training data.
As can be seen, the specification at no time describes obtaining a reliability value for a UE based on a requested metric for training data that includes whether or not the UE has had anomalous behavior related to the training data. As the claim explicitly requires that the request be related to a set of reliability values for a training data set, and comprises a requested metric, the claim requires the anomalous behavior of the user equipment to be related to the training data and that metric, which is not supported by the specification. This causes these limitations to be new matter, and therefore rejected under U.S.C. 112(a).
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
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 1-4, and 6-11, 13-16, and 18 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.
As per claims 1 and 13, this claim calls for “select a subset based on the reliability values for the user equipments and the reliability values for the training data sets.” However, the claim has confusing language when it comes to “the reliability values for the training data sets. The claim calls for “for a respective user equipment, send a request for a set of reliability values for a training data set stored in the respective user equipment.” This seems to call for a set of reliability values for a particular training data set. Then, “obtain for each use equipment in the group, a reliability value, based at least in part on the requested metric, from the set of reliability values for the training data set stored in the user equipment.” This seems to pull a single reliability value from a set of potential reliability values in a particular training set. However, the claim requires the reliability values for the training data sets. Which does not appear to be defined in the claims. There are reliability values for particular training data sets, and a single value pulled for each training data set, but no defined set of reliability values for multiple training data sets. This cause the claim to be confusing, and therefore rejected under U.S.C. 112(b) for failing to particularly point out and claim the intended invention.
As per claims 2-4, 6-11, 14-16 and 18, these claims are rejected as being dependent on a claim rejected under U.S.C. 112(b) for failing to particularly point out and claim the intended invention.
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.
Claims 1-8, and 12-19 are rejected under 35 U.S.C. 103 as being unpatentable over Jha et al (US 20200027022 A1) in view of Hiessl (“Industrial Federated Learning – Requirement and System Design”), Vasseur et al (US 20170279836 A1), and Tuor et al (US 20210158099 A1).
As per claims 1 and 13, Jha discloses, “An apparatus comprising at least one processing core, and at least one memory including computer program code, the at least one memory and the computer program code being configured to, with the at least one processing core, cause the apparatus at least to” (Pg.13-14, particularly paragraph 0139; EN: this denotes the hardware for the system).
“obtain reliability values for each user equipment” (pg.9, particularly paragraph 0096; EN: this denotes using quality of service (i.e. reliability) to determine which participants to select for learning). “in a group of user equipments” (pg.2, particularly paragraph 0020; EN: this denotes the various participant devices that will be part of the system).
“For a respective user equipment in the group, send a request for a set of reliability values… , wherein the request comprises a requested metric” (Pg.9, particularly paragraph 0096; EN: this denotes various metrics from the local participant and a query (i.e. request) for the data).
“obtain, for each user equipment in the group, a … value… for the training data set stored in the user equipment, wherein (i) each user equipment stores a distinct training data set, and …” (Pg.9, particularly paragraph 0096; EN: this denotes looking at the training data for quality values and other information).
“select a subset based on the reliability values for the user equipments and the… values for the training data sets” (pg.9, particularly paragraph 0096; EN: this denotes using quality of service (i.e. reliability) to determine which participants to select for learning).
“direct the subset of the group of user equipment’s to separately perform a machine learning training process in the user equipment’s in the subset” (Pg.4, particularly paragraph 0047; EN: this denotes the various participants performing their machine learning process).
However, Jha fails to explicitly disclose, “send a request for a set of reliability values for a training data set stored in the respective user equipment”, “a reliability value, based at least in part on the requested metric, from the set of reliability values for the training data set”, “and (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior”, and “the reliability values for the training data sets.”
Hiessl discloses, “a reliability value… from the set of reliability values for the training data set” and “the reliability values for the training data sets” (pg.46-47, particularly section 4.3; EN: this denotes looking at Quality of information associated with federated learning clients).
Vasseur discloses, “and (ii) the reliability value indicates whether the respective user equipment is associated with anomalous behavior” (Pg.8-9, particularly paragraph 0090; EN: this denotes monitoring distributed learning agents for interaction with malware related to the Supervisory and control agent related to the distributed learning system).
Tuor discloses, “send a request for a set of reliability values for a training data set stored in the respective user equipment”, “… based at least in part on the requested metric, from the set of reliability values” (Pg.4-5, particularly paragraph 0041; EN: this denotes looking at the local models to determine usefulness metrics of the local datasets of contributors in a federated learning system).
Jha and Hiessl are analogous art because both involve distributed learning.
Before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Hiessl in order to include the reliability of the training data when considering the value of a participant.
The motivation for doing so would be to “decide on the extent of contribution of an individual FL client in the parameter aggregation process” (Hiessl, Pg.46-47, section 4.3) or in the case of Jha, include quality of information when determining participants for their process.
Therefore before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Hiessl in order to include the reliability of the training data when considering the value of a participant.
Jha and Vasseur are analogous art because both involve distributed learning.
Before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Vasseur in order to consider potential anomalous behavior of the devices in a distributed learning system.
The motivation for doing so would be to allow “the anomaly detection infrastructure of the network to be operable to detect network attacks (e.g., DDOs attacks, the sue of malware such as viruses, rootkits, etc)” (Vasseur, Pg.3-4, paragraph 0040) or in the case of Jha, allow the system to consider monitoring edge devices in their distributed learning system in order to avoid those devices using malware and other anomalous behavior situations.
Therefore before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Vasseur in order to consider potential anomalous behavior of the devices in a distributed learning system.
Jha and Tuor are analogous art because both involve federated learning.
Before the effective filing date it would have been obvious to one skilled in the art of federated learning to combine the work of Jha and Tuor in order to include have a set of reliability values for training the data based on the training data.
The motivation for doing so would be to “improve an efficiency with which the federated learning process is performed … the subset of contributors who are currently active or connected need to have datasets that provide a good representation of the entire input space of the global model that is to be trained by the modelling program” (Tuor, Pg.4-5, paragraph 0041) or in the case of Jha, allow the system to evaluate the local data to make sure it is an effective use of the federated learning systems time.
Therefore before the effective filing date it would have been obvious to one skilled in the art of federated learning to combine the work of Jha and Tuor in order to include have a set of reliability values for training the data based on the training data.
As per claims 2 and 14, Jha discloses, “wherein the at least one memory and the computer program code are configured to, with the at least one processing core, cause the apparatus to receive, from each user equipment in the subset, a result of the machine learning training process performed by the user equipment” (Pg.4, particularly paragraph 0047; EN: this denotes the various participants performing their machine learning process).
As per claims 3 and 15, Jha discloses, “wherein the at least one memory and the computer program code are configured to, with the at least one processing core, cause the apparatus to aggregate the results of the machine learning processes received from the user equipments of the subset to obtain an aggregate machine learning result” (Pg.4, particularly paragraph 0047; EN: this denotes the various participants performing their machine learning process and returning the data to the global model to be used).
As per claims 4 and 16, Jha discloses, “wherein the at least one memory and the computer program code are configured to, with the at least one processing core, cause the apparatus to obtain the reliability values for the user equipments in the group from a network data analytics function” (pg.1-2, particularly paragraph 0019; EN: this denotes looking at quality of service such as latency).
As per claims 6 and 18, Jha discloses, “wherein the apparatus is configured to obtain, for each user equipment in the group, from the reliability value of the user equipment and the … value of the training data set stored in the user equipment, a compound reliability value of the user equipment, and to select the subset from among the group based on the compound reliability values of the user equipments of the group” (pg.9-10, particularly paragraphs 0099-0100; EN: this denotes combining the various data in order to select participants).
Hiessl discloses, “reliability values of the training data” (pg.46-47, particularly section 4.3; EN: this denotes looking at Quality of information associated with federated learning clients).
As per claim 7, Jha discloses, “Wherein the apparatus is further configured to store at least one of the compound reliability values of the user equipments of the group in a network node” (pg.14, particularly paragraph 0140; EN: This denotes storing the various code and data being worked with and manipulated).
As per claim 8, Jha discloses, “wherein the network node the apparatus is configured to store the at least one of the compound reliability values of the user equipment of the group in an analytics data repository function” (pg.14, particularly paragraph 0140; EN: This denotes storing the various code and data being worked with and manipulated. This includes network analytics, which meets the broadest reasonable interpretation of an analytics data repository function).
Claim Rejections - 35 USC § 103
Claims 9-10 are rejected under 35 U.S.C. 103 as being unpatentable over Jha et al (US 20200027022 A1) in view of Hiessl (“Industrial Federated Learning – Requirement and System Design”), Vasseur et al (US 20170279836 A1), and Tuor et al (US 20210158099 A1) and further in view of Toy et al (“Overall Network and Service Architecture”).
As per claim 9, Jha discloses, “Wherein the apparatus is configured to receive information from the user equipments comprised in the subset…” (Pg.9, particularly paragraph 0096; EN: this denotes looking at the training data for quality values and other information).
However, Jha fails to explicitly disclose, “using user plane traffic.”
Toy discloses, “using user plane traffic” (pg.107, Sixth paragraph; EN: this denotes using user plane functions for communication over wireless).
Jha and Toy are analogous art because both involve data transmission.
Before the effective filing date it would have been obvious to one skilled in the art of data transmission to combine the work of Jha and Toy in order to make use of user plane traffic.
The motivation for doing so would be because “The 5G system architecture separates the user plane (UP) functions from the control plane (CP) functions, allowing independent scalability, evolution and flexible deployments” (Toy, Pg.108, second to last paragraph) or in the case of Jha, allow the system to use modern 5G components to transfer data effectively and flexibly as needed.
Therefore before the effective filing date it would have been obvious to one skilled in the art of data transmission to combine the work of Jha and Toy in order to make use of user plane traffic.\
As per claim 10, Jha discloses, “wherein the apparatus is configured to receive information from the user equipments comprised in the subset …”
However, Jha fails to explicitly disclose, “using service based architecture signaling or non-access stratum signaling.”
Toy discloses, “using service based architecture signaling or non-access stratum signaling” (pg.107, particularly the fourth paragraph; EN: this denotes 5G being a service-based architecture).
Jha and Toy are analogous art because both involve data transmission.
Before the effective filing date it would have been obvious to one skilled in the art of data transmission to combine the work of Jha and Toy in order to make use of wireless for service based architecture.
The motivation for doing so would be because “The 5G core, as defined by 3GPP, utilized cloud-aligned virtualized functions, service-based architecture (SBA) that spans across all 5 G functions and interactions, including authentication, security, session management, and aggregation of traffic from end devices” (Toy, Pg.107, fourth paragraph) or in the case of Jha, allow the system to use modern 5G components to provide the security, session management and traffic aggregation as needed by the system.
Therefore before the effective filing date it would have been obvious to one skilled in the art of data transmission to combine the work of Jha and Toy in order to make use of wireless for service based architecture.
Claim Rejections - 35 USC § 103
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Jha et al (US 20200027022 A1) in view of Hiessl (“Industrial Federated Learning – Requirement and System Design”), Vasseur et al (US 20170279836 A1), and Tuor et al (US 20210158099 A1) and further in view of Hirose (US 20070189322 A1).
As per claim 11, Jha fails to explicitly disclose, “wherein the apparatus is configured to notify user equipments comprised in the group but not comprised in the subset, that they have been excluded from the machine learning training process.”
Hirose discloses, “wherein the apparatus is configured to notify user equipments comprised in the group but not comprised in the subset, that they have been excluded from the machine learning training process” (abstract; EN: this denotes notifying devices that they have not been selected so they don’t have to take further action of setting parameters).
Jha and Hirose are analogous art because both involve data transmission.
Before the effective filing date it would have been obvious to one skilled in the art of data transmission to combine the work of Jha and Hirose in order to notify devices they have not been selected.
The motivation for doing so would be so that “the communication parameters are not set in unintended pair of devices” or in the case of Jha, allow the system to notify client devices that they don’t need to perform any further actions when they are not selected for processing.
Therefore before the effective filing date it would have been obvious to one skilled in the art of data transmission to combine the work of Jha and Hirose in order to notify devices they have not been selected.
Claim Rejections - 35 USC § 103
Claims 12 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Jha et al (US 20200027022 A1) in view of Hiessl (“Industrial Federated Learning – Requirement and System Design”) and Vasseur et al (US 20170279836 A1).
As per claims 12 and 19, Jha discloses, “An apparatus comprising at least one processing core, at least one memory including computer program code, the at least one memory and the computer program code being configured to, with the at least one processing core, cause the apparatus at least to:” (Pg.13-14, particularly paragraph 0139; EN: this denotes the hardware for the system).
“store a training data locally in the apparatus” (Pg.9, particularly paragraph 0096; EN: this denotes looking at the training data for quality values and other information associated with local devices).
“Provide, responsive to a request, comprising a requested metric from a federated learning server” (pg.9, particularly paragraph 0096; EN: this denotes using quality of service (i.e. reliability) to determine which participants to select for learning along with other metrics such as training data quality, training data size, AoI, etc). “a … value for the set of training data to the federated learning server” (Pg.9, particularly paragraph 0096; EN: this denotes looking at the training data for quality values and other information).
“perform a machine learning training process using the set of training data as a response to an instruction from the federated learning server” (Pg.4, particularly paragraph 0047; EN: this denotes the various participants performing their machine learning process).
However, Jha fails to explicitly disclose, “a reliability value for the set of training data, wherein the reliability value is based at least in part on the requested metric” and ”(ii) indicates whether the apparatus is associated with anomalous behavior.”
Hiessl discloses, “a reliability value for the set of training data, wherein the reliability value is based at least in part on the requested metric” (pg.46-47, particularly section 4.3; EN: this denotes looking at Quality of information associated with federated learning clients).
Vasseur discloses, ”(ii) indicates whether the apparatus is associated with anomalous behavior” (Pg.8-9, particularly paragraph 0090; EN: this denotes monitoring distributed learning agents for interaction with malware related to the Supervisory and control agent related to the distributed learning system).
Jha and Hiessl are analogous art because both involve distributed learning.
Before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Hiessl in order to include the reliability of the training data when considering the value of a participant.
The motivation for doing so would be to “decide on the extent of contribution of an individual FL client in the parameter aggregation process” (Hiessl, Pg.46-47, section 4.3) or in the case of Jha, include quality of information when determining participants for their process.
Therefore before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Hiessl in order to include the reliability of the training data when considering the value of a participant.
Jha and Vasseur are analogous art because both involve distributed learning.
Before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Vasseur in order to consider potential anomalous behavior of the devices in a distributed learning system.
The motivation for doing so would be to allow “the anomaly detection infrastructure of the network to be operable to detect network attacks (e.g., DDOs attacks, the sue of malware such as viruses, rootkits, etc)” (Vasseur, Pg.3-4, paragraph 0040) or in the case of Jha, allow the system to consider monitoring edge devices in their distributed learning system in order to avoid those devices using malware and other anomalous behavior situations.
Therefore before the effective filing date it would have been obvious to one skilled in the art of distributed learning to combine the work of Jha and Vasseur in order to consider potential anomalous behavior of the devices in a distributed learning system.
Response to Arguments
In pg.10, the Applicant argues in regards to the rejection under U.S.C. 101 of independent claims 1 and 13,
Regarding Prong One, the Office Action alleges that "The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion." Office Action, page 2. The MPEP does define the mental processes grouping as "concepts performed in the human mind (including observation, evaluation, judgment, opinion)." MPEP §2106.04(a)(2)(II). It further notes that "[c]laims do not recite a mental process when they do not contain limitations that can practically be performed in the human mind, for instance when the human mind is not equipped to perform the claim limitation." Id. Independent claims 1 and 13 recite an enhanced process for federated machine learning by selecting only the most reliable devices and data, based on whether a reliability value indicates that a particular user equipment is associated with anomalous behavior. For example, independent claims 1 and 13 recite, in some form or another, to:
In response, the Examiner maintains the rejection as shown above. Claims 1 and 13 describe the mental process of selecting user equipment to use in training a system based upon their reliability and any anomalous behavior, which is a mental process as described above. This not an improvement to federated machine learning because this is selecting data to be put into the model, not an improvement or change to the model itself. This causes the improvement to be to the abstract idea and not the machine learning process or the hardware of the system, therefore the rejection is maintained as shown above.
In pg.11, the Applicant further argues in regards to the rejection under U.S.C. 101,
None of the above emphasized operations may be practically performed within the human mind. The human mind, for example, cannot practically obtain, for each user equipment in the group, a reliability value. Nor can the human mind practically direct a subset of the group of user equipments to separately perform a machine learning training process in the user equipments in the subset. The claims are not directed to a judicial exception under prong one of Step 2A. For at least these reasons, Applicant respectfully requests withdrawal of the rejection under 35 U.S.C. § 101.
In response, the Examiner maintains the rejection as shown above. The transmission of data between systems (i.e. the obtaining steps) and the use of generic machine learning/training are not mental processes, but examples of using generic computer systems/generic machine learning models to “apply” the abstract idea. Here the claims amount to no more than the passing of reliability data between generic computer systems to provide information for the abstract idea of selecting reliable training data for the generic machine learning model. Since these portions are not designated as mental processes, they are not enough to provide significantly more than the abstract idea, and therefore the rejection is maintained as shown above.
In pg.12, the Applicant further argues in regards to the rejection under U.S.C. 101),
Even if the claims were hypothetically considered to be directed to an abstract idea of mental processes, independent claims 1 and 13 recite additional elements that integrate any abstract idea into a practical application because the claims include specific features that are specifically designed to achieve an improved technological result for distributed and federated machine learning in wireless communication networks. For example, certain embodiments of the present application improve the performance and robustness of distributed or federated learning systems. The embodiments described by the present application enhance distributed machine learning training by integrating reliability evaluation for both participating nodes and their local training data. Instead of randomly selecting devices to contribute updates, the system computes reliability values based on factors such as UE behavior, software integrity, network conditions, and the completeness and relevance of collected data. By using these reliability values to select only the most dependable nodes and data sources for each training round, the described approach significantly reduces the influence of faulty, malicious, or low-quality contributors. This targeted participation results in faster convergence and higher accuracy, yielding a clear technical improvement. The described improvements are evident in the claims by the recitations shown above.
In response, the Examiner maintains the rejection as shown above. As discussed above, the aspects of the claims the Applicant discusses are no more than the passing of data between generic computer systems and the use of generic machine learning models. The claims give no details of the machine learning model/federated learning system other than the name of the well known algorithm. Merely manipulating the data going into a generic computer system performing generic machine learning is not an improvement to the hardware or the machine learning model, it is an improvement to the abstract idea of selecting appropriate data for the machine learning system. Therefore the rejection is maintained as shown above.
In pg.13, Applicant argues in regards to the rejection under U.S.C. 101,
Applicant refers to BASCOM Global Internet Servs. V. AT& TMobility LLC, 827F.3d 1341, 1350, 119 USPQ2d 1236, 1242 (Fed. Cir. 2016), which holds that an "inventive concept may be found in the non-conventional and non-generic arrangement of components that are individually well-known and conventional." MPEP 2106.05 further indicates, regarding Bascom, "The district court should have considered the additional elements in combination, because the "inventive concept inquiry requires more than recognizing that each claim element, by itself, was known in the art" (827 F.3d at 1350, 119 USPQ2d at 1242)." In Bascom, individual claim elements were found to be non-conventional. However, the courts found the non-generic arrangement of components that are individually well-known to provide the patent eligible matter. Similarly, in the present case, even if one were to consider that individual elements are routine and conventional, the arrangement of all the claim features provides a non-conventional improvement to distributed and federated machine learning in wireless communication networks. Applicant submits that in view of Bascom, the claims are patent eligible.
In response, the Examiner maintains the rejection as shown above. There is no non-conventional or non-generic arrangement of components here. The claims call for the transference of information between generic “user equipment” to “an apparatus.” In federated learning, the system always includes multiple edge devices in communication with one another. There is nothing unconventional or non-generic in the hardware or machine learning model, and therefore the BASCOM elements do not apply, and the rejection is maintained as shown above.
Applicant's arguments with respect to claims 1-4, 6-16, and 18-19 have been considered but are either moot in view of the new ground(s) of rejection or repetitions of the above arguments and rejected for similar reasons given above.
Conclusion
The examiner requests, in response to this Office action, support be shown for language added to any original claims on amendment and any new claims. That is, indicate support for newly added claim language by specifically pointing to page(s) and line no(s) in the specification and/or drawing figure(s). This will assist the examiner in prosecuting the application.
When responding to this office action, Applicant is advised to clearly point out the patentable novelty which he or she thinks the claims present, in view of the state of the art disclosed by the references cited or the objections made. He or she must also show how the amendments avoid such references or objections See 37 CFR 1.111(c).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BEN M RIFKIN whose telephone number is (571)272-9768. The examiner can normally be reached Monday-Friday 9 am - 5 pm.
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, Alexey Shmatov can be reached at (571) 270-3428. 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.
/BEN M RIFKIN/Primary Examiner, Art Unit 2123
Prosecution Insights