Detailed Action
Claims 1-20 are presently pending.
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 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 1-10 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 1, the claim recites “based on the transmitting, receiving, from the server, information relating to training of the second instance of the AI model.” It is unclear which device, the cloud system or the device, receives the information relating to training of the second instance of the AI model.
For purpose of the examination, the limitation is interpreted to mean: The device receives the information relating to training of the second instance of the AI model.
Claim 1 recites “receiving, from the quantum computer, at least one output of the one or more computations;” It is unclear which device receives the output from the quantum computer.
For purpose of the examination, the limitation is interpreted to mean: The server receives the output of the quantum computer.
Claims 2-10 depend from the apparatus claim 1. Therefore, claims 2-10 inherit the same deficiency.
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 1-4, 6-8 and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Kim al. (US 20240177039 A1, hereinafter ‘Kim’) in view of Stapleton et al. (US 11399017 B1, hereinafter ‘Stapleton’) and further in view of Zhang et al. (US 20220044162 A1, hereinafter ‘Zhang’).
Regarding claim 1, Kim teaches:
A device, comprising: a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, the operations comprising: ([0033]-[0034] discloses that the server and the devices are implemented using generic computer components including operating systems)
[Fig. 2] and [0035]-[0036] The Global QSNN in the global server is the first instance of the AI model, and the Local QSNN is the second instance of the AI model)
based on the transmitting, receiving, from the server, information relating to training of the second instance of the AI model; ([Kim, Fig. 6, S10-S20] The server 100 transmits the initialized neural network model to at least one local device 200. The local device 200 trains it by inputting observation value [Fig. 6, S30])
[Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. [0117] The quantum slimmable neural network model 330 includes encoder unit 310 and a quantum circuit unit 330 hardwares)
receiving, from the quantum computer, at least one output of the one or more computations; ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
providing the at least one output to the server, wherein the providing enables the server to aggregate the at least one output with one or more other outputs provided by one or more other devices involved in the FL; ([Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer)
obtaining aggregated data from the server based on the providing; and ([Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer. [Fig. 6, S60] and [0122] Global server retransmits global side parameters to at least one local device)
utilizing the aggregated data to update the second instance of the AI model. ([Fig. 6, S60] and [0122] and [0124] Global server retransmits global side parameters to at least one local device to update local models)
However, Kim does not specifically disclose:
transmitting, to a server, a request to participate in federated learning (FL) for an artificial intelligence (AI) model,
after the receiving, determining that one or more computations associated with the training are to be performed by a quantum computer;
based on the determining, causing at least a portion of a local dataset accessible to the device to be provided to the quantum computer to perform the one or more computations;
Stapleton teaches:
after the receiving, determining that one or more computations [Stapleton, Fig. 1A] The Classical Computer (CC) 101 is the server, according to [col 6, lines 28-43]. Quantum Computer (QC) 104, Simulated Quantum Computer (SQC) 106, and Client Device 102 are edge nodes connected to the server.
[Stapleton, col 10, lines 26-48] The encryption circuit 222A in the CC 101 receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data to the quantum computer 104 or simulated quantum computer 106 for decryption)
based on the determining, causing at least a portion of a local dataset accessible to the device to be provided to the quantum computer to perform the one or more computations; ([Stapleton, col 10, lines 26-48] The encryption circuit receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data (local data) to the quantum computer for decryption (computations))
receiving, from the quantum computer, at least one output of the one or more computations; ([Stapleton, col 10, lines 26-48] The encryption circuit receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data to the quantum computer for decryption. The quantum computer 104 transmit the decrypted (computed) data to a device. The Classical Computer (CC) 101 is the server, according to [col 6, lines 28-43], and [col 12, lines 16-26; Fig. 1A] the CC 101 is coupled to communication network 120 allowing the CC to receive decrypted data 114 from SQC or QC)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim and Stapleton to use the method of selecting whether to use quantum computer or a classical computer of Stapleton to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the efficiency of the learning system. Quantum Computing device, Simulated Quantum Computing device, and conventional processors have performance difference, therefore anyone who knows the art would introduce the task distribution method based on the amounts of resources available to improve the efficiency of the system [Stapleton, col 7, lines 7-24] and [Stapleton, col 10, lines 26-48].
However, Kim in view of Stapleton does not specifically disclose:
transmitting, to a server, a request to participate in federated learning (FL) for an artificial intelligence (AI) model.
Zhang teaches:
transmitting, to a server, a request to participate in federated learning (FL) for an artificial intelligence (AI) model. ([0042; Fig. 2] The clients 204 send requests to participate in training the machine learning model at operations 214. In response to the receipt of the request, the central server send global machine-learning model data216 to clients)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton and Zhang to use the method of verify authenticity of the device using a software defined network system of Zhang to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the efficiency of the training system by activating the training system only when it is needed. Training of a machine learning model requires significant computing resources, and performing it only when it is needed helps the system to conserve resources.
Regarding claim 2, Kim teaches:
The device of claim 1, wherein the aggregated data are also utilized to update the first instance of the AI model in the cloud system. ([Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer. [Fig. 6, S60] and [0122] Global server retransmits global side parameters to at least one local device)
Regarding claim 3, Kim teaches:
The device of claim 1, wherein the device comprises a quantum node. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
Regarding claim 4, Kim teaches:
The device of claim 1, wherein the information comprises AI model configuration data, one or more AI model data structures, one or more AI model parameters, data regarding an AI model sharing state, or a combination thereof. ([Kim, Fig. 6, S10-S20] The server 100 transmits the initialized neural network model (i.e., configuration data) to at least one local device 200. The local device 200 trains it by inputting observation value [Fig. 6, S30])
Regarding claim 6, Kim teaches:
The device of claim 1, wherein no portion of the local dataset is provided from the device to the cloud system. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The observation value is not transferred to the server)
Regarding claim 7, Kim teaches:
The device of claim 1, wherein the local dataset comprises raw training data for the AI model. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation)
Regarding claim 8, Kim in view of Stapleton teaches:
The device of claim 1, wherein the determining comprises determining that the one or more computations require more than a threshold amount of computational resources. ([Stapleton, col 10, lines 26-48] The encryption circuit receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data (local data) to the quantum computer for decryption (computations). The quantum computer 104 transmit the decrypted (computed) data back to a computing device)
Regarding claim 10, Kim in view of Stapleton and further in view of Zhang teaches:
The device of claim 1, wherein the transmitting is performed in response to a receipt of a service request by the device. ([0042; Fig. 2] The clients 204 send requests to participate in training the machine learning model at operations 214. In response to the receipt of the request, the central server send global machine-learning model data216 to clients)
Claims 5, 9, and 11-20 are rejected under 35 U.S.C. 103 as being unpatentable over Kim in view of Stapleton in view of Zhang and further in view of Mehmedagic et al. (US 20210029029 A1, hereinafter ‘Mehmedagic’)
Regarding claim 5, Kim in view of Stapleton in view of Zhang teaches:
The device of claim 1.
Kim in view of Stapleton in view of Zhang does not specifically disclose:
wherein the server comprises a fog server.
Mehmedagic teaches:
wherein the server comprises a fog server. ([Mehmedagic, Fig. 3; 0072 and 0074] The compute nodes 342 are connected to the fog server controller 335)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton, Zhang and Mehmedagic to use the method of using a fog server of Mehmedagic to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the integrity and efficiency of the system. The fog computing method of Mehmedagic allows the system to reduce latency and save network bandwidth by processing at least some portion of computations at the computer closer to the edge nodes [Mehmedagic, 0065 and 0164].
Regarding claim 9, Kim in view of Stapleton in view of Zhang teaches:
The device of claim 1.
Kim in view of Stapleton in view of Zhang does not specifically disclose:
wherein the transmitting triggers the server to access a software defined network (SDN) system to verify authenticity of the device.
Mehmedagic teaches:
wherein the transmitting triggers the server to access a software defined network (SDN) system to verify authenticity of the device. ([Mehmedagic, 0263] When a network device receives a new message and the network has no rules for handling a message from the new device, SDN is used to determine whether the new industrial device is authorized)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton, Zhang and Mehmedagic to use the method of verify authenticity of the device using a software defined network system of Mehmedagic to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the integrity of the system by preventing unauthorized intrusion or a cyberattack [Mehmedagic, 0227 and 0263].
Regarding claim 11, Kim teaches:
A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising: ([0033]-[0034] discloses that the server and the devices are implemented using generic computer components including operating systems)
[Fig. 2] and [0035]-[0036] The Global QSNN in the global server is the first instance of the AI model, and the Local QSNN is the second instance of the AI model)
based on the obtaining, transmitting, to the device, information relating to training of the second instance of the AI model, at least a portion of a local dataset accessible to the device to be provided to the quantum computer to perform the one or more computations[Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. [0117] The quantum slimmable neural network model 330 includes encoder unit 310 and a quantum circuit unit 330 hardwares)
after the transmitting, receiving the at least one output from the device; ([Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model)
aggregating the at least one output with one or more other outputs provided by one or more other devices involved in the FL, resulting in aggregated data; and ([Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer. [Fig. 6, S60] and [0122] Global server retransmits global side parameters to at least one local device)
sending the aggregated data to the device, thereby enabling the device to update the second instance of the AI model. ([Fig. 6, S60] and [0122] and [0124] Global server retransmits global side parameters to at least one local device to update local models)
However, Kim does not specifically disclose:
receiving, from a device, a request to participate in federated learning (FL) for an artificial intelligence (AI) model
based on the receiving, accessing a software defined network (SDN) controller to verify authenticity of the device;
responsive to the accessing, obtaining an indication from the SDN controller that the device has been verified;
wherein the transmitting causes the device to determine whether one or more computations associated with the training are to be performed by a quantum computer, cause at least a portion of a local dataset accessible to the device to be provided to the quantum computer to perform the one or more computations based on a determination that the one or more computations are to be performed by the quantum computer, and receive at least one output of the one or more computations from the quantum computer;
Stapleton teaches:
wherein the transmitting causes the device to determine whether one or more computations [Stapleton, Fig. 1A] The Classical Computer (CC) 101 is the server, according to [col 6, lines 28-43]. Quantum Computer (QC) 104, Simulated Quantum Computer (SQC) 106, and Client Device 102 are edge nodes connected to the server.
[Stapleton, col 10, lines 26-48] The encryption circuit receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data to the quantum computer for decryption. The quantum computer 104 transmit the decrypted (computed) data to a device. The Classical Computer (CC) 101 is the server, according to [col 6, lines 28-43], and [col 12, lines 16-26; Fig. 1A] the CC 101 is coupled to communication network 120 allowing the CC to receive decrypted data 114 from SQC or QC)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Zhang and Stapleton to use the method of selecting whether to use quantum computer or a classical computer of Stapleton to implement the federated learning system of Zhang. The suggestion and/or motivation to do so is to improve the efficiency of the learning system. Quantum Computing device, Simulated Quantum Computing device, and conventional processors have performance difference, therefore anyone who knows the art would introduce the task distribution method based on the amounts of resources available to improve the efficiency of the system [Stapleton, col 7, lines 7-24] and [Stapleton, col 10, lines 26-48].
However, Kim in view of Stapleton does not specifically disclose:
receiving, from a device, a request to participate in federated learning (FL) for an artificial intelligence (AI) model
based on the receiving, accessing a software defined network (SDN) controller to verify authenticity of the device;
responsive to the accessing, obtaining an indication from the SDN controller that the device has been verified;
Zhang teaches:
receiving, from a device, a request to participate in federated learning (FL) for an artificial intelligence (AI) model ([Zhang, 0042; Fig. 2] The clients 204 send requests to participate in training the machine learning model at operations 214. [0032] The Central Server 202 and Blockchain 206 are interpreted as the cloud system, as the definition of the cloud is ‘data is processed and accessed via the internet, rather than on a local server’ and the global model stored in the central server and blockchain are accessed via network. [0028] The “local machine-learning model” is the second instance of the AI model, and the global machine-learning model stored in the central server is the first instance of the AI model)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton and Zhang to use the method of verify authenticity of the device using a software defined network system of Zhang to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the efficiency of the training system by activating the training system only when it is needed. Training of a machine learning model requires significant computing resources, and performing it only when it is needed helps the system to conserve resources.
However, Kim in view of Stapleton in view of Zhang does not specifically disclose:
based on the receiving, accessing a software defined network (SDN) controller to verify authenticity of the device;
responsive to the accessing, obtaining an indication from the SDN controller that the device has been verified;
Mehmedagic teaches:
based on the receiving, accessing a software defined network (SDN) controller to verify authenticity of the device; ([Mehmedagic, 0263] When a network device receives a new message and the network has no rules for handling a message from the new device, SDN is used to determine whether the new industrial device is authorized)
responsive to the accessing, obtaining an indication from the SDN controller that the device has been verified; ([Mehmedagic, 0263-0264; Fig. 15] When a network device receives a new message and the network has no rules for handling a message from the new device, SDN is used to determine whether the new industrial device is authorized. In response to the authorization, a network path (indication) is provided)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton, Zhang and Mehmedagic to use the method of verify authenticity of the device using a software defined network system of Mehmedagic to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the integrity of the system by preventing unauthorized intrusion or a cyberattack [Mehmedagic, 0227 and 0263].
Regarding claim 12, Kim in view of Stapleton in view of Zhang and further in view of Mehmedagic teaches:
The non-transitory machine-readable medium of claim 11, wherein the processing system is implemented in a fog server. ([Mehmedagic, Fig. 3; 0072 and 0074] The compute nodes 342 are connected to the fog server controller 335)
Regarding claim 13, Kim teaches:
The non-transitory machine-readable medium of claim 11, wherein the device comprises a quantum node. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
Regarding claim 14, Kim teaches:
The non-transitory machine-readable medium of claim 11, wherein the local dataset comprises raw training data for the AI model. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
Regarding claim 15, Kim in view of Stapleton and further in view of Zhang teaches:
The non-transitory machine-readable medium of claim 11, wherein the device and the one or more other devices comprise a subset of devices selected by the processing system for participating in the FL. ([0051] At operations 310, the clients 304 read the metadata of the other clients and determine which of the client devices (the subset of devices) will be participating in a given training round)
Regarding claim 16, Kim teaches:
A method, comprising: [Fig. 2] and [0035]-[0036] The Global QSNN in the global server is the first instance of the AI model, and the Local QSNN is the second instance of the AI model)
based on the transmitting, receiving, by the processing system, [Kim, Fig. 6, S10-S20] The server 100 transmits the initialized neural network model to at least one local device 200. The local device 200 trains it by inputting observation value [Fig. 6, S30])
[Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
receiving, by the processing system, at least one computational output from the quantum computer associated with the at least some of the training; ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
providing, by the processing system, and to the[Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer. [Fig. 6, S60] and [0122] Global server retransmits global side parameters to at least one local device)
obtaining, by the processing system, aggregated data from the [Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer. [Fig. 6, S60] and [0122] Global server retransmits global side parameters to at least one local device)
updating, by the processing system, the second instance of the AI model based on the aggregated data. ([Fig. 6, S60] and [0122] and [0124] Global server retransmits global side parameters to at least one local device to update local models)
However, Kim does not specifically disclose:
transmitting, by a processing system of a device including a process, and to a fog server, a request to participate in federated learning (FL) for an artificial intelligence (AI) model,
receiving, … from the fog server, information
after the receiving, determining, by the processing system, that resources of a quantum computer are needed for at least some of the training;
based on the determining, causing, by the processing system, at least a portion of a local dataset to be provided to the quantum computer;
receiving, by the processing system, at least one computational output from the quantum computer associated with the at least some of the training;
providing, … to the fog server, the at least one computational output
obtaining, … data from the fog server
Stapleton teaches:
after the receiving, determining, by the processing system, that resources of a quantum computer are needed for (computation) [Stapleton, Fig. 1A] The Classical Computer (CC) 101 is the server, according to [col 6, lines 28-43]. Quantum Computer (QC) 104, Simulated Quantum Computer (SQC) 106, and Client Device 102 are edge nodes connected to the server.
[Stapleton, col 10, lines 26-48] The encryption circuit 222A in the CC 101 receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data to the quantum computer 104 or simulated quantum computer 106 for decryption)
based on the determining, causing, by the processing system, at least a portion of a local dataset to be provided to the quantum computer; ([Stapleton, col 10, lines 26-48] The encryption circuit receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data (local dataset) to the quantum computer for decryption. The quantum computer 104 transmit the decrypted (computed) data back to a computing device)
receiving, by the processing system, at least one computational output from the quantum computer [Stapleton, col 10, lines 26-48] The encryption circuit receives the resource parameter, determine whether the processing time exceeds the predetermined threshold value, and if it exceeds the threshold, sends the data to the quantum computer for decryption. The quantum computer 104 transmit the decrypted (computed) data to a device. The Classical Computer (CC) 101 is the server, according to [col 6, lines 28-43], and [col 12, lines 16-26; Fig. 1A] the CC 101 is coupled to communication network 120 allowing the CC to receive decrypted data 114 from SQC or QC)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim and Stapleton to use the method of selecting whether to use quantum computer or a classical computer of Stapleton to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the efficiency of the learning system. Quantum Computing device, Simulated Quantum Computing device, and conventional processors have performance difference, therefore anyone who knows the art would introduce the task distribution method based on the amounts of resources available to improve the efficiency of the system [Stapleton, col 7, lines 7-24] and [Stapleton, col 10, lines 26-48].
However, Kim in view of Stapleton does not specifically disclose:
transmitting, by a processing system of a device including a process, and to a fog server, a request to participate in federated learning (FL) for an artificial intelligence (AI) model,
receiving, … from the fog server, information
providing, … to the fog server, the at least one computational output
obtaining, … data from the fog server
Zhang teaches:
transmitting, by a processing system of a device including a process, and to a [Zhang, 0042; Fig. 2] The clients 204 send requests to participate in training the machine learning model at operations 214. [0032] The Central Server 202 and Blockchain 206 are interpreted as the cloud system, as the definition of the cloud is ‘data is processed and accessed via the internet, rather than on a local server’ and the global model stored in the central server and blockchain are accessed via network. [0028] The “local machine-learning model” is the second instance of the AI model, and the global machine-learning model stored in the central server is the first instance of the AI model)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton and Zhang to use the method of verify authenticity of the device using a software defined network system of Zhang to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the efficiency of the training system by activating the training system only when it is needed. Training of a machine learning model requires significant computing resources, and performing it only when it is needed helps the system to conserve resources.
However, Kim in view of Stapleton and further in view of Zhang does not specifically disclose:
receiving, … from the fog server, information
providing, … to the fog server, the at least one computational output
obtaining, … data from the fog server
Mehmedagic teaches:
receiving, … from the fog server, information ([0088] The fog controller 435 provides control inputs to the server)
providing, … to the fog server, the at least one computational output ([0074], [0077] and [0088] The computes nodes 342 are managed by the fog controller 335 and provide the compute resources to the system. The fog controller 435 provides control inputs to the server)
obtaining, … data from the fog server ([0088] The fog controller 435 provides control inputs to the server)
Before the effective filing date of the invention to a person of ordinary skill in the art, it would have been obvious, having the teachings of Kim, Stapleton, Zhang and Mehmedagic to use the method of using a fog server of Mehmedagic to implement the federated learning system of Kim. The suggestion and/or motivation to do so is to improve the integrity and efficiency of the system. The fog computing method of Mehmedagic allows the system to reduce latency and save network bandwidth by processing at least some portion of computations at the computer closer to the edge nodes [Mehmedagic, 0065 and 0164].
Regarding claim 17, Kim teaches:
The method of claim 16, wherein the aggregated data are also utilized to update the first instance of the AI model in the cloud system. ([Fig. 6, S40] and [0120]-[0121] Global server combines parameters received from at least one local devices to train the global side quantum neural network model. The parameter from the local device comes from the quantum circuit 330 which is the quantum computer. [Fig. 6, S60] and [0122] Global server retransmits global side parameters to at least one local device)
Regarding claim 18, Kim teaches:
The method of claim 16, wherein the processing system is implemented in a quantum node. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The generated parameters are transmitted to global server side)
Regarding claim 19, Kim teaches:
The method of claim 16, wherein no portion of the local dataset is provided from the device to the cloud system. ([Fig. 6, S30] and [0116] The local device inputs at least one observation value (i.e., local dataset) into the quantum slimmable neural network model which includes quantum circuits 330 to perform training computation. The observation value is not transferred to the server)
Regarding claim 20, Kim in view of Stapleton in view of Zhang and further in view of Mehmedagic teaches:
The method of claim 16, wherein the transmitting triggers the fog server to access a software defined network (SDN) system to verify authenticity of the device. ([Mehmedagic, 0159] shows that the authentication service may reside on a physical compute node of the fog server. [0263-0264; Fig. 15] When a network device receives a new message and the network has no rules for handling a message from the new device, SDN is used to determine whether the new industrial device is authorized. In response to the authorization, a network path (indication) is provided)
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JUN KWON whose telephone number is (571)272-2072. The examiner can normally be reached Monday – Friday 8:00AM – 5:00PM ET.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Abdullah Kawsar can be reached at (571)270-3169. 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.
/JUN KWON/Examiner, Art Unit 2127
/TEWODROS E MENGISTU/Examiner, Art Unit 2127
Prosecution Insights