DETAILED ACTION
This action is in response to the amendments filed 5/1/2026. Claims 1, 3, 5-14, and 16-20 are pending and have been examined.
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 .
Priority
Acknowledgment is made of applicant's claim for foreign priority based on an application filed in the People’s Republic of China on 8/25/2021. It is noted, however, that applicant has not filed a certified copy of the App #CN202110981600.0 application as required by 37 CFR 1.55.
Should applicant desire to obtain the benefit of foreign priority under 35 U.S.C. 119(a)-(d) prior to declaration of an interference, a certified English translation of the foreign application must be submitted in reply to this action. 37 CFR 41.154(b) and 41.202(e).
Failure to provide a certified translation may result in no benefit being accorded for the non-English application.
Claim Interpretation
The following is a quotation of MPEP 2111.04 II:
The broadest reasonable interpretation of a method (or process) claim having contingent limitations requires only those steps that must be performed and does not include steps that are not required to be performed because the condition(s) precedent are not met. For example, assume a method claim requires step A if a first condition happens and step B if a second condition happens. If the claimed invention may be practiced without either the first or second condition happening, then neither step A or B is required by the broadest reasonable interpretation of the claim. If the claimed invention requires the first condition to occur, then the broadest reasonable interpretation of the claim requires step A. If the claimed invention requires both the first and second conditions to occur, then the broadest reasonable interpretation of the claim requires both steps A and B.
The broadest reasonable interpretation of a system (or apparatus or product) claim having structure that performs a function, which only needs to occur if a condition precedent is met, requires structure for performing the function should the condition occur. The system claim interpretation differs from a method claim interpretation because the claimed structure must be present in the system regardless of whether the condition is met and the function is actually performed.
Limitation 7 of Claim 1 (“in response to the time consumption of the alternative combination satisfying a schedulable constraint parameter, determining the alternative combination as the candidate combination”) recites a step of determining a candidate combination, only performed if the time consumption of the alternative combination satisfies a schedulable constraint parameter. Thus, this limitation is found to be contingent, and is consequently interpreted as not being a required component of the claimed method under broadest reasonable interpretation. To become a required limitation of the method, it must be rewritten as a positively recited element.
Claim Objections
Applicant is advised that should claim 14 be found allowable, claim 16 will be objected to under 37 CFR 1.75 as being a substantial duplicate thereof. When two claims in an application are duplicates or else are so close in content that they both cover the same thing, despite a slight difference in wording, it is proper after allowing one claim to object to the other as being a substantial duplicate of the allowed claim. See MPEP § 608.01(m).
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, 3, 5-14, and 16-20 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.
Claim 1 recites “wherein allocating the N first tasks to the K network models for operation … comprises” before its sixth limitation and “the method further comprises” before its eighth limitation. It’s unclear whether “wherein allocating the N first tasks to the K network models for operation … comprises” is meant to encompass all following limitations, or just the two limitations preceding “the method further comprises”. Thus, the scope of the claim is rendered indefinite. This deficiency is present in substantially similar independent claim 14, and inherited by all dependent claims. “wherein allocating the N first tasks to the K network models for operation … comprises” is interpreted as encompassing the following two limitations, not including limitations following “the method further comprises”.
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, 3, 5-14, and 16-20 are rejected under 35 U.S.C. 101 because the claimed inventions are directed to non-statutory subject matter without significantly more.
Claim 1
Step 1: The claim recites “A multi-task deployment method”, and is therefore directed to the statutory category of process
Step 2A Prong 1: The claim recites the following judicial exception(s)
selecting a target combination with a maximum combination operation accuracy from the at least one candidate combination: This can be performed as a mental process. One can merely observe the target combination with maximum accuracy.
in response to the time consumption of the alternative combination satisfying a schedulable constraint parameter, determining the alternative combination as the candidate combination: This can be performed as a mental process. One can merely observe an alternative combination satisfying a schedulable constraint parameter.
determining a total number of iterations based on N and K: This can be performed as a mental process. One can merely decide on a total number of iterations based on N and K.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
obtaining N first tasks and K network models, wherein N and K are positive integers greater than or equal to 1: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
allocating the N first tasks to the K network models for operation, to obtain at least one candidate combination of tasks and network models, wherein each candidate combination comprises a mapping relation between the N first tasks and the K network models: This is mere instruction to obtain at least one candidate combination based on tasks and models in a generic manner (MPEP 2106.05(f)).
deploying a target mapping relation and the K network models on a prediction machine, wherein the target mapping relation is the mapping relation in the target combination, and the prediction machine is a device that directly performs prediction, and the device is able to predict a task through a deployed network model, and output a prediction result: This is mere instruction to deploy a target mapping relation on a generic computer (MPEP 2106.05(f)).
wherein allocating the N first tasks to the K network models for operation, to obtain the at least one candidate combination of the tasks and the network models, comprises: in response to the N first tasks being allocated, obtaining a time consumption of task execution of an alternative combination of the tasks and the network models obtained by allocating of the N first tasks: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination: This is mere instruction to find a combination with PSO based on combination operation accuracy and in response to an iteration threshold in a generic manner (MPEP 2106.05(f)).
wherein the PSO algorithm uses all alternative combinations as particles, a fitness value of each particle is obtained based on the combination operation accuracy of the alternative combination, a global optimal position (Pbest) and a global extreme value (Gbest) are updated according to the fitness value, and a position and a speed of the particle is also updated, and it is determined whether Gbest reaches a maximum number of iterations or whether the Pbest satisfies a minimum limit, so as to filter the alternative combinations: This is mere instruction to calculate a fitness value based on combination operator accuracy in a generic manner and filter alternative combinations in a generic manner (MPEP 2106.05(f)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
obtaining N first tasks and K network models, wherein N and K are positive integers greater than or equal to 1: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
allocating the N first tasks to the K network models for operation, to obtain at least one candidate combination of tasks and network models, wherein each candidate combination comprises a mapping relation between the N first tasks and the K network models: This is mere instruction to obtain at least one candidate combination based on tasks and models in a generic manner (MPEP 2106.05(f)).
deploying a target mapping relation and the K network models on a prediction machine, wherein the target mapping relation is the mapping relation in the target combination, and the prediction machine is a device that directly performs prediction, and the device is able to predict a task through a deployed network model, and output a prediction result: This is mere instruction to deploy a target mapping relation on a generic computer (MPEP 2106.05(f)).
wherein allocating the N first tasks to the K network models for operation, to obtain the at least one candidate combination of the tasks and the network models, comprises: in response to the N first tasks being allocated, obtaining a time consumption of task execution of an alternative combination of the tasks and the network models obtained by allocating of the N first tasks: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination: This is mere instruction to find a combination with PSO based on combination operation accuracy and in response to an iteration threshold in a generic manner (MPEP 2106.05(f)).
wherein the PSO algorithm uses all alternative combinations as particles, a fitness value of each particle is obtained based on the combination operation accuracy of the alternative combination, a global optimal position (Pbest) and a global extreme value (Gbest) are updated according to the fitness value, and a position and a speed of the particle is also updated, and it is determined whether Gbest reaches a maximum number of iterations or whether the Pbest satisfies a minimum limit, so as to filter the alternative combinations: This is mere instruction to calculate a fitness value based on combination operator accuracy in a generic manner and filter alternative combinations in a generic manner (MPEP 2106.05(f)).
Claim 3
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites no further judicial exception(s)
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
in response to the time consumption of the alternative combination not satisfying the schedulable constraint parameter, discarding the alternative combination and obtaining a next alternative combination: This is a routine limitation for a neural architecture search and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
in response to the time consumption of the alternative combination not satisfying the schedulable constraint parameter, discarding the alternative combination and obtaining a next alternative combination: This is an instance of searching through candidate models based on a performance metric, a widely used technique and boilerplate function of neural architecture searches, as noted by Nagaraja (SYSTEM AND METHOD OF CREATING ARTIFICIAL INTELLIGENCE MODEL, MACHINE LEARNING MODEL OR QUANTUM MODEL GENERATION FRAMEWORK, filed 9/18/2020, US 20210334700 A1): “Neural architecture search (NAS) is a technique for automating the design of artificial neural networks (ANN), a widely used model in the field of machine learning … NAS finds an architecture from all possible architectures by following a search strategy that will maximize the performance and typically includes three dimensions a) a search space, b) a search strategy and c) a performance estimation. The search space is an architecture pattern that is typically designed by an NAS approach. The search strategy is something that depends upon the search methods used to define a NAS approach, for example a Bayesian optimization or a reinforcement learning. The search strategy accounts for the time taken to build a model. The performance estimation is the convergence of certain performance metrics expected out of a NAS produced neural architecture model” (Nagaraja, [0003]).
Claim 5
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining the time consumption of the alternative combination based on the present WCET of each first task and a present task processing cycle: This can be performed as a mental process. One can merely sum the WCETs across all first asks and divide the total by the task processing cycle to obtain a relative time consumption of the alternative combination.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
obtaining a present Worst Case Execution Time (WCET) of each first task of the N first tasks in the alternative combination when the first task is executed on an assigned target network model: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
obtaining a present Worst Case Execution Time (WCET) of each first task of the N first tasks in the alternative combination when the first task is executed on an assigned target network model: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II.)
Claim 6
Step 1: The claim recites a process, as in claim 5
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining a total WCET of the alternative combination based on the present WCET of each first task: This can be performed as a mental process. One can simply sum the present WCET across all first tasks for the alternative combination.
obtaining the time consumption of the alternative combination based on the total WCET of the alternative combination and the present task processing cycle: This can be performed as a mental process. One can merely divide the total WCET by the present task processing cycle to get a relative measure of time consumption for the alternative combination.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 7
Step 1: The claim recites a process, as in claim 6
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining an average WCET of the first task on the target network model based on the plurality of historical WCETs and the present WCET: This can be performed as a mental process. One can merely calculate an average of all the historical and present WCETs.
obtaining the total WCET of the alternative combination based on the average WCET of each first task: This can be performed as a mental process. One can merely assign the average WCET as the total WCET.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
for each first task, obtaining a plurality of historical WCETs of the target network model corresponding to the first task: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
for each first task, obtaining a plurality of historical WCETs of the target network model corresponding to the first task: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II.)
Claim 8
Step 1: The claim recites a process, as in claim 7
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining a first standard deviation of the plurality of historical WCETs and the present WCET: This can be performed as a mental process. One can merely calculate the standard deviation of the historical and present WCETs.
obtaining a first sum value of the average WCET and the first standard deviation: This can be performed as a mental process. One can merely add the first sum to the average WCET and the standard deviation.
obtaining the total WCET of the alternative combination by summing the first sum value of each first task in the alternative combination: This can be performed as a mental process. One can merely add the first sum value across all first tasks for the alternative combination.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 9
Step 1: The claim recites a process, as in claim 6
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining an average task processing cycle based on the plurality of historical task processing cycles and the present task processing cycle: This can be performed as a mental process. One can merely calculate an average of the historical and present task processing cycle.
determining the time consumption of the alternative combination based on the total WCET and the average task processing cycle: This can be performed as a mental process. One can merely divide the total WCET by the average task processing cycle to determine a relative time consumption of the alternative combination.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
obtaining a plurality of historical task processing cycles: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
obtaining a plurality of historical task processing cycles: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II.)
Claim 10
Step 1: The claim recites a process, as in claim 9
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining a standard deviation of the plurality of historical task processing cycles and the present task processing cycle: This can be performed as a mental process. One can merely calculate the standard deviation of the historical and present task processing cycles.
obtaining a sum value of the average task processing cycle and the second standard deviation: This can be performed as a mental process. One can merely sum the average task processing cycle value and the second standard deviation.
obtaining a ratio of the total WCET to the sum value as the time consumption of the alternative combination: This can be performed as a mental process. One can merely divide the total WCET by the second sum value.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 11
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining a combination operation accuracy of the candidate combination based on the task operation accuracy of each first task in the candidate combination: This can be performed as a mental process. One can merely sum the task operation accuracies across all first tasks of the candidate combination.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
for each candidate combination, obtaining a task operation accuracy of each first task of the N first tasks in the candidate combination executed on the assigned target network model: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
for each candidate combination, obtaining a task operation accuracy of each first task of the N first tasks in the candidate combination executed on the assigned target network model: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II.)
Claim 12
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
obtaining the combination operation accuracy of the candidate combination by weighting the task operation accuracy of each first task based on the weight of each first task: This can be performed as a mental process. One can merely calculate a weighted sum of task operation accuracies across all first asks.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
obtaining a weight of each first task: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
obtaining a weight of each first task: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II.)
Claim 13
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites no further judicial exception(s)
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
in response to receiving a second task within a target task processing cycle, sorting second tasks to be processed within the target task processing cycle: This is mere instruction to sort tasks in a generic manner (MPEP 2106.05(f)).
querying the target mapping relation for the second tasks in sequence to obtain a target network model corresponding to the currently queried second task: This is mere instruction to obtain a target network model from a target mapping relation in a generic manner (MPEP 2106.05(f)).
issuing the currently queried second task to the target network model on the prediction machine for processing: This is mere instruction to issue a task to a model on a generic computer (MPEP 2106.05(f)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
in response to receiving a second task within a target task processing cycle, sorting second tasks to be processed within the target task processing cycle: This is mere instruction to sort tasks in a generic manner (MPEP 2106.05(f)).
querying the target mapping relation for the second tasks in sequence to obtain a target network model corresponding to the currently queried second task: This is mere instruction to obtain a target network model from a target mapping relation in a generic manner (MPEP 2106.05(f)).
issuing the currently queried second task to the target network model on the prediction machine for processing: This is mere instruction to issue a task to a model on a generic computer (MPEP 2106.05(f)).
Claim 14
Step 1: The claim recites “An electronic device”, and is therefore directed to the statutory category of article of manufacture
Step 2A Prong 1: The claim recites the following judicial exception(s)
selecting a target combination with a maximum combination operation accuracy from the at least one candidate combination: This can be performed as a mental process. One can merely observe the target combination with maximum accuracy.
in response to the time consumption of the alternative combination satisfying a schedulable constraint parameter, determining the alternative combination as the candidate combination: This can be performed as a mental process. One can merely observe an alternative combination satisfying a schedulable constraint parameter.
determining a total number of iterations based on N and K: This can be performed as a mental process. One can merely decide on a total number of iterations based on N and K.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
at least one processor; and a memory communicatively coupled to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, when the instructions are executed by the at least one processor, the at least one processor is enabled to implement the following operations: This is mere instruction to execute the recited judicial exception(s) on generic computer hardware (MPEP 2106.05(f)).
obtaining N first tasks and K network models, wherein N and K are positive integers greater than or equal to 1: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
allocating the N first tasks to the K network models for operation, to obtain at least one candidate combination of tasks and network models, wherein each candidate combination comprises a mapping relation between the N first tasks and the K network models: This is mere instruction to obtain at least one candidate combination based on tasks and models in a generic manner (MPEP 2106.05(f)).
deploying a target mapping relation and the K network models on a prediction machine, wherein the target mapping relation is the mapping relation in the target combination, and the prediction machine is a device that directly performs prediction, and the device is able to predict a task through a deployed network model, and output a prediction result: This is mere instruction to deploy a target mapping relation on a generic computer (MPEP 2106.05(f)).
wherein allocating the N first tasks to the K network models for operation, to obtain the at least one candidate combination of the tasks and the network models, comprises: in response to the N first tasks being allocated, obtaining a time consumption of task execution of an alternative combination of the tasks and the network models obtained by allocating of the N first tasks: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination: This is mere instruction to find a combination with PSO based on combination operation accuracy and in response to an iteration threshold in a generic manner (MPEP 2106.05(f)).
wherein the PSO algorithm uses all alternative combinations as particles, a fitness value of each particle is obtained based on the combination operation accuracy of the alternative combination, a global optimal position (Pbest) and a global extreme value (Gbest) are updated according to the fitness value, and a position and a speed of the particle is also updated, and it is determined whether Gbest reaches a maximum number of iterations or whether the Pbest satisfies a minimum limit, so as to filter the alternative combinations: This is mere instruction to calculate a fitness value based on combination operator accuracy in a generic manner and filter alternative combinations in a generic manner (MPEP 2106.05(f)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
at least one processor; and a memory communicatively coupled to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, when the instructions are executed by the at least one processor, the at least one processor is enabled to implement the following operations: This is mere instruction to execute the recited judicial exception(s) on generic computer hardware (MPEP 2106.05(f)).
obtaining N first tasks and K network models, wherein N and K are positive integers greater than or equal to 1: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II.)
allocating the N first tasks to the K network models for operation, to obtain at least one candidate combination of tasks and network models, wherein each candidate combination comprises a mapping relation between the N first tasks and the K network models: This is mere instruction to obtain at least one candidate combination based on tasks and models in a generic manner (MPEP 2106.05(f)).
deploying a target mapping relation and the K network models on a prediction machine, wherein the target mapping relation is the mapping relation in the target combination, and the prediction machine is a device that directly performs prediction, and the device is able to predict a task through a deployed network model, and output a prediction result: This is mere instruction to deploy a target mapping relation on a generic computer (MPEP 2106.05(f)).
wherein allocating the N first tasks to the K network models for operation, to obtain the at least one candidate combination of the tasks and the network models, comprises: in response to the N first tasks being allocated, obtaining a time consumption of task execution of an alternative combination of the tasks and the network models obtained by allocating of the N first tasks: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination: This is mere instruction to find a combination with PSO based on combination operation accuracy and in response to an iteration threshold in a generic manner (MPEP 2106.05(f)).
wherein the PSO algorithm uses all alternative combinations as particles, a fitness value of each particle is obtained based on the combination operation accuracy of the alternative combination, a global optimal position (Pbest) and a global extreme value (Gbest) are updated according to the fitness value, and a position and a speed of the particle is also updated, and it is determined whether Gbest reaches a maximum number of iterations or whether the Pbest satisfies a minimum limit, so as to filter the alternative combinations: This is mere instruction to calculate a fitness value based on combination operator accuracy in a generic manner and filter alternative combinations in a generic manner (MPEP 2106.05(f)).
Claim 16
Step 1: The claim recites an article of manufacture, as in claim 14
Step 2A Prong 1: The claim recites the following further judicial exception(s)
determining a total number of iterations based on N and K: This can be performed as a mental process. One can merely calculate the maximum number of combinations of models.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination: This is mere instruction to find a combination with PSO based on combination operation accuracy and in response to an iteration threshold in a generic manner (MPEP 2106.05(f)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination: This is mere instruction to find a combination with PSO based on combination operation accuracy and in response to an iteration threshold in a generic manner (MPEP 2106.05(f)).
Claims 17-19
Step 1: Claims 17-19 recite an article of manufacture, as in claim 14.
Step 2A Prong 1: Claims 17-19 recite the same judicial exception(s) as claims 5, 11, and 13, respectively.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through any additional elements. The analysis of claims 17-19 at this step mirrors that of claims 5, 11, and 13, respectively, with the exception that claims 17-19 are directed to “at least one processor; and 4507547 a memory communicatively coupled to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, when the instructions are executed by the at least one processor, the at least one processor is enabled to implement the following operations”, said operations mirroring those of claims 5, 11, and 13. This is a mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s). The analysis of claims 17-19 at this step mirrors that of claims 5, 11, and 13, with the exception that claims 17-19 are directed to “at least one processor; and 4507547 a memory communicatively coupled to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, when the instructions are executed by the at least one processor, the at least one processor is enabled to implement the following operations”, said operations mirroring those of claims 5, 11, and 13. This is mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Claim 20
Step 1: The claim recites “A non-transitory computer-readable storage medium”, and is therefore directed to the statutory category of article of manufacture
Step 2A Prong 1: The claim recites the judicial exception(s) of claim 1
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
A non-transitory computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are configured to cause a computer to implement the method of claim 1: This is mere instruction to execute the recited judicial exception(s) with generic computer hardware (MPEP 2106.05(f)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
A non-transitory computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are configured to cause a computer to implement the method of claim 1: This is mere instruction to execute the recited judicial exception(s) with generic computer hardware (MPEP 2106.05(f)).
Allowable Subject Matter
Claims 1, 3, 5-14, and 16-20 would be allowable over the prior art of record if rewritten or amended to overcome the rejection(s) under 35 U.S.C. 101, 35 U.S.C. 112(b), set forth in this Office action.
Regarding claim 1, “determining a total number of iterations based on N and K; and in response to the total number of iterations being greater than an iteration number threshold, searching for a next alternative combination through a Particle Swarm Optimization (PSO) algorithm based on a combination operation accuracy of the alternative combination” is not taught by the prior art of record. The closest prior arts of record are B. Wang et al. (Evolving Deep Neural Networks by Multi-objective Particle Swarm Optimization for Image Classification, published 4/22/2019, arXiv:1904.09035v2, retrieved from https://arxiv.org/abs/1904.09035), hereafter referred to as ‘B. Wang’, M. Wang et al. (DISTRIBUTED FAILOVER FOR MULTI-TENANT SERVER FARMS, published 5/5/2016, US 20160124818 A1), hereafter referred to as ‘M. Wang’, and La Lumondiere et al. (OPTIMIZED ILLUMINATION FOR IMAGING, published 5/1/2014, US 20140118561 A1), hereafter referred to as ‘La Lumondiere’.
B. Wang discloses…
Using particle swarm optimization (PSO) to perform a neural architecture search: “This paper proposes a novel multi-objective optimization method for evolving state-of-the-art deep CNNs in real-life applications, which automatically evolves the non-dominant solutions at the Pareto front. Three major contributions are made: Firstly, a new encoding strategy is designed to encode one of the best state-of-the-art CNNs; With the classification accuracy and the number of floating point operations as the two objectives, a multi-objective particle swarm optimization method is developed to evolve the non-dominant solutions” (B. Wang, page 1, left column, abstract)
… that optimizes models based on a combination accuracy of the alternative combinations: “The overall goal of this paper is to propose a multi-objective particle swarm optimization (MOPSO) method to balance the trade-off between the classification accuracy and the inference latency (time), which is named MOCNN.” (B. Wang, page 2, left column, paragraph 2)
B. Wang does not disclose determining a number of iterations based on a number of tasks and a number of models or using PSO conditionally based on an iteration threshold.
M. Wang discloses dynamically executing either an enumerative algorithm or a greedy algorithm to process server-client pairings based on the number of possible server-client combinations:
“In other implementations, additional or alternative algorithms may be used. For example, for a relatively small number of machine and small number of users, an enumeration algorithm may be used instead of the greedy algorithm. In a simplified example, for 5 servers and 3 users on each server, all possible backup combinations may be enumerated, and the one with the minimum failover time may be chosen. The number of combinations can be calculated in this way: for the 3 users in the first machine, each of them will choose a machine as backup, so the number of combinations is 81. The same calculation can be used for users in the other 4 machines. Thus, the total number of combinations is 405.” (M. Wang, [0090])
“For a large number of severs and a large number of users, e.g., 1000 servers and 50 users on each sever, the total combinations would be 1000*(999̂50)=9.512*E152, which is a huge number of combinations and is not able to be enumerated in an acceptable time. In this case, the greedy algorithm may be used” (M. Wang, [0091])
“In order to choose between the enumerated algorithm and the greedy algorithm, the equations used to calculate the total number of combinations may be utilized. Specifically, as long as the number of machine and the number of users are given, the total number of combinations can be calculated. If an available computer can enumerate, e.g., one million combinations in several seconds, a threshold may be configured to choose between the enumeration and greedy algorithms, based on an available/allowable time to execute the algorithms.” (M. Wang, [0092])
M. Wang does not disclose determining a number of iterations based on a number of tasks and a number of network models, using particle swarm optimization, or searching for a combination based on the accuracy of a combination.
La Lumondiere discloses executing particle swarm optimization in response to a number of iterations of a previous genetic algorithm being reached: “in some embodiments, a GSA algorithm may be and/or utilize a hybrid algorithm having aspects of a genetic algorithm and a particle swarm optimizer algorithm. Such hybrid algorithms may retain positive aspects of both genetic algorithms and particle swarm optimizer algorithms and may lead to efficiently determined optimal solutions regardless of problem structure. In one example embodiment, the computer 1208 may run a genetic algorithm for a threshold number of generations (iterations) and/or until a fitness function value or values reaches a predetermined level, at which point a particle swarm optimization algorithm is used until convergence is reached.” (La Lumondiere, [0068])
La Lumondiere does not disclose determining a total number of iterations based on a number of models and a number of tasks or executing PSO based on operation accuracy of alternative combinations.
Therefore, the prior art of record, individually or in combination, does not disclose the entirety of claim 1 as a whole. Substantially similar independent claim 14 is allowable under 35 U.S.C. 102 and 103 under this rationale. Claims 3, 5-13, and 16-20 are allowable at least due to their dependence on claims 1 and 14.
Regarding claim 9, the limitation “determining the time consumption of the alternative combination based on the total WCET and the average task processing cycle”, in combination with the other limitations of the claim, is not taught by the prior art of record. The closest prior arts of record are Bird et al. (ADAPTIVE RESOURCE USAGE LIMITS FOR WORKLOAD MANAGEMENT, published 6/19/2014, US 20140173616 A1), hereafter referred to as ‘Bird’, and Ajmera et al. (EFFICIENT MECHANISM FOR EXECUTING SOFTWARE-BASED SWITCHING PROGRAMS ON HETEROGENOUS MULTICORE PROCESSORS, PCT filed 12/24/2018, US 2022/0012108 A1), hereafter referred to as ‘Ajmera’.
Bird discloses converting application execution time into a proportionate processor utilization metric by dividing processor time by the total processor time available to the system, potentially using historical metrics collected over some number of time intervals: “The set of scheduling tasks performed each interval according to an embodiment of the present invention is illustrated in FIG. 8. At step 810, the scheduler thread collects metrics on the aggregate processor time or processor utilization (e.g., percent of processor capacity) used by all threads doing work on behalf of a particular application or workload since the last collection period. In the case where processor time is the only metric available, the thread will convert the processor time used by each individual workload into a processor utilization metric by dividing it by the total processor time available on the system over that collection interval. The exact calculation will be operating system and/or hardware dependent, and also dependent on whether the system is running under a virtualized environment. Alternatively, the processor utilization may be computed based on metrics collected over the last N intervals, rather than only the most recent interval, as a way to have the system adapt more smoothly to changes in workload. In such cases a moving average may be used to determine the processor utilization compared to the activation threshold discussed below.” (Bird, [0041])
Bird does not disclose measuring worst-case execution time as a metric for processor time.
Ajmera discloses calculating an average task processing cycle based on an obtained plurality of historical task processing cycles: “The dynamic thread assignment component 150 may then determine a value that is indicative of the number of active processing cycles used by that polling thread 145. The value may be, for example, an average number of active processing cycles per second used by the polling thread 145 over a period of time (T). In one embodiment, the value is a cumulative (historical) average that averages the measurements taken over several intervals.” (Ajmera, [0033]).
Ajmera does not disclose determining time consumption based on worst case execution time(s) and the average task processing cycle.
Therefore, the prior art of record, individually or in combination, does not disclose the entirety of claim 9 as a whole. Claim 10 would be allowable at least due to its dependence on claim 9.
Response to Arguments
The following responses address arguments and remarks made in the instant remarks dated 5/1/2026.
Priority
A certified copy of foreign application CN202110981600.0 has not been filed. Thus, benefit of foreign priority under 35 U.S.C. 119(a)-(d) has not been obtained.
Claim Interpretation
The Examiner notes that one limitation of claim 1 is found to be contingent, and is thus interpreted as not being a requirement of the claimed method under broadest reasonable interpretation. See the claim interpretation section for more detail.
Objections
Previous objections to the specification have been withdrawn in light of the instant amendments.
The Applicant is warned that duplicate claims 14 and 16 will result in an objection upon one being found allowable. See the claim objections section for more detail.
112 Rejections
Previous rejections under 35 U.S.C. 112(b) have been withdrawn in light of the instant amendments.
However, new rejections under 35 U.S.C. 112(b) have been made in light of the instant amendments.
101 Rejections
On page 15 of the instant remarks, the Applicant argues that the claims do not recite abstract ideas:
“Independent claims 1, and 14 as amended clearly recite elements such as "tasks",
"network models", "PSO algorithm", and therefore are not a "mental process" as contended by the
Office Action nor are the elements "a method for organizing human activity." Therefore, Applicant
disagrees that the claims are directed to an abstract idea under Step 2A, Prong One.”
In regards to the Applicant’s arguments above, the Examiner respectfully disagrees that the claims, as amended, recite no mental processes. Claims scrutinized under 35 U.S.C. 101 are analyzed on a per-limitation basis to identify the recitation of abstract ideas, as noted in MPEP 2106.04(a): Examiners should determine whether a claim recites an abstract idea by (1) identifying the specific limitation(s) in the claim under examination that the examiner believes recites an abstract idea, and (2) determining whether the identified limitations(s) fall within at least one of the groupings of abstract ideas listed above.
Claim 1 recites limitations amounting to mentally performable processes (such as limitation 3: “selecting a target combination with a maximum combination operation accuracy from the at least one candidate combination”). The presence of additional elements that cannot be performed as mental processes in other limitations does not render all limitations non-abstract. Similar reasoning is applicable to all other claims.
The Examiner asserts that the claims, as amended, recite mental processes, and maintains their rejections on the basis of the Alice/Mayo tests performed (See 101 rejections section for more detail).
On pages 15-20 of the instant remarks, the Applicant argues that recited judicial exceptions are integrated into a practical application by improving on existing technology, and the claimed invention amounts to significantly more on the basis of said improvements:
“Assuming, arguendo, that the claims are directed to one of the subject matter
groupings of abstract ideas enumerated under step 2A, Prong One, which Applicant does not concede,
the concepts of the rejected claims are integrated into a practical application such that the claims are
not directed to an abstract idea. Applicant respectfully asserts that the claims are clearly a practical
application, in that the claims generate and produce a tangible result that provides a meaningful
improvement over existing technology as recognized by the specification. See, e.g., MPEP § 2106.05.
…
Here, the amended claims recite a technical solution as the claims are directed to a
multi-task deployment method, electronic device, and non-transitory computer-readable storage
medium. For example, in independent claims 1, and 14, after the target mapping relation included in
the target combination and the K network models are deployed on the prediction machine, in response
to receiving a first task, the corresponding network model in the K network models can be called
based on the target mapping relation, and the operation on the first task can be performed by the
corresponding network model. By matching the tasks with the network models, the optimal
combination of tasks and network models can be obtained, thereby improving the timeliness
and accuracy of task processing. Further, in some cases where the amount of data is relatively large,
the PSO algorithm can be used to filter the alternative combinations, thereby reducing the
amount of operation data and reducing the cost. Furthermore, in amended claims 1, and 14, the
target mapping relation included in the target combination and the K network models are deployed
on a prediction machine which is used to perform prediction.
…
With the multi-task deployment method according to embodiments of the disclosure,
problems in related arts that the usage time and the model chosen to run a certain task are not precisely
designed (as it is difficult to match a deep learning model to a task empirically to ensure real-time schedulability) can be solved. By obtaining N first tasks and K network models, allocating the N first
tasks to the K network models for operation to obtain at least one candidate combination of tasks and
network models, selecting a target combination with a maximum combination operation accuracy
from the at least one candidate combination, deploying a target mapping relation and the K network
models on a prediction machine, the optimal combination of tasks and network models can be
obtained, thereby improving the timeliness and accuracy of task processing. In addition, by using the
PSO algorithm to filter the alternative combinations, the amount of operation data and the cost can be
reduced.”
In response to the Applicant’s argument that the claimed invention improves upon existing technology, the Examiner respectfully disagrees. The improvement of a claimed invention must be sufficiently detailed, as noted in MPEP 2106.05(a): “If it is asserted that the invention improves upon conventional functioning of a computer, or upon conventional technology or technological processes, a technical explanation as to how to implement the invention should be present in the specification. That is, the disclosure must provide sufficient details such that one of ordinary skill in the art would recognize the claimed invention as providing an improvement. The specification need not explicitly set forth the improvement, but it must describe the invention such that the improvement would be apparent to one of ordinary skill in the art … After the examiner has consulted the specification and determined that the disclosed invention improves technology, the claim must be evaluated to ensure the claim itself reflects the disclosed improvement in technology. Intellectual Ventures I LLC v. Symantec Corp., 838 F.3d 1307, 1316, 120 USPQ2d 1353, 1359 (Fed. Cir. 2016) (patent owner argued that the claimed email filtering system improved technology by shrinking the protection gap and mooting the volume problem, but the court disagreed because the claims themselves did not have any limitations that addressed these issues). That is, the claim must include the components or steps of the invention that provide the improvement described in the specification.”
It’s not clear to the Examiner what technical field or existing technology is purportedly improved by the claimed invention, or what advantages particle swarm optimization offers over existing conventional techniques for filtering candidate architectures. Selection of an optimal task-architecture pairing is a standard, baseline practice for the entire fields of automated machine learning and neural architecture searching. The Applicant must detail what issue exists in an existing technology or technical field that is addressed by the claimed invention, cite where these improvements are described in the instant specification, and ensure aspects of the invention required for said improvements are represented in the claim language.
Thus, no rejections are withdrawn on this basis. See the 101 rejections section for more detail.
102 / 103 Rejections
On pages 20-21 of the instant remarks, the Applicant argues that subject matter previously indicated as allowable has been brought into the independent claims, and thus all claims should be allowable under 35 U.S.C. 102 and 103:
“Applicant has amended claim 1 by including the technical features of claim 4, and
amended claim 14 by including the technical features of claim 16. Thus the amended claims 1 and 14
should be allowable. When independent claims 1 and 14 are patentable over the prior art, claims
dependent therefrom are patentable. Applicant therefore respectfully requests that the rejections of
claims 1, and 14 and the claims that depend thereon under 35 U.S.C. § 102 & § 103 be withdrawn.”
The Applicant’s arguments above have been fully considered and are persuasive. Previous rejections under 35 U.S.C. 102 and 103 have been withdrawn.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Ramezani et al. (Task-Based System Load Balancing in Cloud Computing Using Particle Swarm Optimization, published 2014, Int J Parallel Prog (2014) 42:739–754) discloses a method of using particle swarm optimization to schedule tasks in a load-balancing model
Wilhelm et al. (The Worst-Case Execution Time Problem - Overview of Methods and Survey of Tools, published 5/8/2008, ACM Transactions on Embedded Computing Systems (TECS), Volume 7, Issue 3, pp. 1-53) discloses that WCETs serve as upper bounds on program performance, which are needed to show that performance requirements of a real-time system are being satisfied.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Aaron P Gormley whose telephone number is (571)272-1372. The examiner can normally be reached Monday - Friday 12:00 PM - 8:00 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 T 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.
/AG/Examiner, Art Unit 2148
/MICHELLE T BECHTOLD/Supervisory Patent Examiner, Art Unit 2148
Prosecution Insights