A lightweight federated learning aggregation method based on genetic algorithm

Abstract

The present invention relates to the field of knowledge graph technology, and in particular to a lightweight federated learning aggregation method based on a genetic algorithm. The present invention provides a lightweight federated learning aggregation method based on a genetic algorithm. By introducing genetic learning theory, the aggregation mechanism of the server side in the federated learning process is reconstructed, and the existing aggregation method is lightweight improved to better solve the problems that arise when the existing technology is applied to the field of the Internet of Things. The present invention significantly accelerates the convergence process of the global model, reduces resource consumption during training, and improves the final performance of the model by selecting client models that perform well in each round of training to participate in aggregation. Compared with existing federated learning algorithms, the selective aggregation scheme of the present invention can effectively reduce or avoid interference from malicious clients when dealing with unreliable or heterogeneous clients, reduce the negative impact of malicious attacks on the performance of federated learning, and further improve the robustness and reliability of the model.

Description

Technical Field

[0001] The present invention relates to the technical field of federated learning, and in particular to a lightweight federated learning aggregation method based on genetic algorithm. Background Art

[0002] With the continuous development of the Internet of Things (IoT), various devices in our lives are becoming increasingly intelligent. These devices can exchange data over the internet, enabling data sharing and resource integration. These interconnected devices are called the Internet of Things (IoT). IoT devices generate enormous amounts of data, known as big data. Big data refers to data that cannot be effectively processed and analyzed by traditional data processing technologies due to its scale, velocity, and complexity. Big data has the following characteristics: 1. Large scale: Big data can reach terabytes or even petabytes; 2. Rapid velocity: Big data is generated at a very high rate, requiring real-time processing; 3. High complexity: Big data contains a wide variety of information types and structures.

[0003] In the context of IoT big data, federated learning has gained widespread application. Federated learning is a distributed machine learning technique that allows multiple parties to collaboratively train models while maintaining data privacy and security. The core advantage of this technique lies in data localization: data does not need to leave its original storage location, thus reducing the risk of data leakage and meeting the requirements of privacy protection laws and regulations. The traditional federated learning framework involves the following three steps: 1. The server randomly selects clients participating in the current round of aggregation and sends the global model to them. 2. Clients that receive the global model update their own models and perform several rounds of training to obtain new local models, which they send back to the server. 3. The server aggregates all received local models to obtain a new global model.

[0004] Federated learning aggregation algorithms. The aggregation process is a crucial step in the federated learning framework, occurring on the server side. The quality of the aggregation algorithm directly impacts training results. Within the federated learning field, aggregation processes primarily fall into two categories: synchronous and asynchronous.

[0005] Asynchronous federated learning allows the client to asynchronously upload locally updated model parameters, and the server asynchronously aggregates these parameters to form a global model. The main purpose of this asynchronous update is to prevent the server from waiting for clients with slow training speeds, causing the training process to become abnormally slow. In other words, this method can speed up the training speed of federated learning to a certain extent and enhance the overall flexibility and scalability of the system. However, this may lead to the problem of using outdated models to update the global model. To solve this problem, researchers have proposed a variety of strategies, such as discarding outdated models, dynamically adjusting the number of local training rounds, and performing time-weighted aggregation updates on the model. However, this phenomenon can only be alleviated, not completely solved, which will have a relatively large impact on the performance of the final global model. Therefore, the present invention chooses to improve the federated learning framework from the perspective of non-federated learning.

[0006] Synchronous federated learning requires that clients receive a sufficient number of client models before performing a round of model aggregation, thus creating the problem of waiting for slow clients. To address this issue, numerous researchers have explored solutions to accelerate training while maintaining model performance by introducing more efficient update mechanisms. FedPA's design requires that the server only wait for a pre-specified number of device models, specified by the agent, in each round before using these models for aggregation. Similar considerations led to the development of an adaptive selection mechanism to determine the time threshold and the client queue for participating in training iterations. However, both approaches ignore the value of data from low-performing clients, which can negatively impact model performance. Other researchers believe that better models can be obtained by selecting more appropriate participating clients or by using specific strategies to filter the models aggregated in each round. Eiffel assigns a priority index based on client performance, and the scheduling algorithm uses this priority as a weight to randomly select participating clients. MS selects models for aggregation based on momentum, selecting only clients whose Euclidean distance to at least half of the other clients is within a certain range for averaging. FedAU adaptively uses a total weight inversely proportional to the average client participation rate, thereby improving model accuracy. The aforementioned scheme of filtering and aggregating clients based on ratings is highly unfair to some clients. Furthermore, some external attackers may attack some key clients, resulting in a decrease in overall training effectiveness. Furthermore, in the evaluation-based selection scheme, the number of client models screened in each round is fixed and close to each other, which may result in the loss of some valid client data, leading to slow training.

[0007] A genetic algorithm is a search and optimization method based on the principles of natural selection and genetics. In a genetic algorithm, feasible solutions to a problem are typically encoded as "genes" represented by one-dimensional vectors. Initially, the program generates a large number of random genotypes as the initial population. The "environment" then assigns a fitness score to each solution in the population. Based on the scores, the best solutions are selected and participate in a crossover and mutation process, forming a new population. This process is repeated until a set of high-quality solutions with fitness scores that meet a certain standard is obtained. Summary of the Invention

[0008] The purpose of the present invention is to solve the problems of slow training speed, high resource consumption and susceptibility to extreme customers when federated learning is applied to the field of Internet of Things in the prior art, and to propose a lightweight federated learning aggregation method based on genetic algorithm.

[0009] On the one hand, the present invention proposes a lightweight federated learning aggregation method based on a genetic algorithm, characterized by including a server side, the server side specifically including a communication interface module, a model selection module and a model aggregation module, and the server side implements aggregation of the global model of the Internet of Things by comprising the following steps:

[0010] S1. The server randomly selects a set amount of IoT device data from the global verification set to supplement the data of the client that sent the data supplement request;

[0011] S2. The server implements the client selection function through the communication interface module and distributes the current global model of the Internet of Things device to the client.

[0012] S3, the server receives the global model returned after training by the client through the communication interface module

[0013] S4, the server side selects the global model through the model selection module Fitness scoring is performed based on the performance in the global validation set;

[0014] S5: The server selects the top ρ in performance ranking based on the fitness score using a binary heap data structure. t the global model of

[0015] S6. The server aggregates the selected global models to form a new global model through the model aggregation module.

[0016] Preferably, the lightweight genetic algorithm-based federated learning aggregation method further requires looping steps S2 to S7, and ends the loop when the global model reaches a set accuracy or the total number of communication rounds is exhausted.

[0017] Preferably, the server needs to wait for a fixed set time t after receiving the data supplement request sent by the client. wait After completing the data transmission, the server needs to wait for a fixed set time t wait .

[0018] Preferably, the server randomly selects a set amount of IoT device data from the global verification set to supplement data for the client that sends the data supplement request, and this only happens once.

[0019] Preferably, the global validation set is an introduced public data set, and the global validation set provides a model evaluation benchmark for the client training in step S4; the global validation set also serves as a benchmark for fitness scoring in step S5.

[0020] Further preferably, the global validation set needs to satisfy such that the server can fairly characterize the client's environment; the global validation set also needs to satisfy such that the server can evaluate the model returned by the client based on the fitness scoring benchmark established by the global validation set, and select the model participating in the aggregation based on the fitness scoring criteria.

[0021] Preferably, the scores obtained in step S5 are ranked in the top t The global model is added to the new set middle, and The following relationship exists:

[0022]

[0023] Where: Represents the set of clients selected to participate in this communication round.

[0024] Preferably, the new global model in step S6 uses the new set Perform aggregation as follows:

[0025]

[0026] Where: represents the new global model; The global model returned by the server according to the client; represents the new set; p irepresents the aggregate weight of the model from client i.

[0027] Preferably, the number of the global models participating in the aggregation in step S6 is p t , ρ t It can be set according to different situations, as follows:

[0028] The first setting method is to maintain the number of aggregations at a constant value, denoted as:

[0029] ρ t =ρ max

[0030] Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregates during the experiment;

[0031] The second setting method uses a power function, and the aggregate quantity change process can be expressed as:

[0032] ρ t =ρ max *(1-b t )+1while b∈(0,1)

[0033] Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregations during the experiment; b is the base number that needs to be set before the experiment;

[0034] The third setting method is based on a linearly changing number of aggregations, as follows:

[0035]

[0036] Where: ρ t is the number of the global models participating in the aggregation; ρ max represents the maximum value of the number of aggregations during the experiment; c represents the maximum number of iterations of the maximum change in the number of aggregations;

[0037] The fourth design method is based on the sinusoidal variation of the aggregate quantity. The first sinusoidal variation of the aggregate quantity model follows a period of The sine function is written as:

[0038]

[0039] Another model uses a sine function with a complete π cycle, which is written as:

[0040]

[0041] Where: ρt is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregates during the experiment.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] The present invention provides a lightweight federated learning aggregation method based on genetic algorithms. By introducing genetic learning theory, the aggregation mechanism of the server side in the federated learning process is reconstructed, and the existing aggregation method is lightweight improved by using a binary heap data structure and applied to the field of the Internet of Things. On the basis of ensuring communication efficiency and device-side computing power, the training of the distributed learning mechanism optimization model is realized. The present invention significantly accelerates the convergence process of the global model, reduces resource consumption during training, and improves the final performance of the model by selecting client models that perform well in each round of training to participate in aggregation. Compared with existing federated learning algorithms, the selective aggregation scheme of the present invention can effectively reduce or avoid interference from malicious clients when dealing with unreliable or heterogeneous clients, reduce the negative impact of malicious attacks on the performance of federated learning, and further improve the robustness and reliability of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a block diagram of a lightweight genetic algorithm-based federated learning aggregation method in an embodiment of the present invention.

[0045] Figure 2 These are experimental results obtained on different data sets in the embodiments of the present invention.

[0046] Figure 3 These are the experimental results of different numbers of clients on the CIFAR-10 dataset in an embodiment of the present invention.

[0047] Figure 4 These are the experimental results of different Dirichlet parameter scenarios on the CIFAR-10 dataset in an embodiment of the present invention.

[0048] Figure 5 These are the experimental results under different attack modes in the embodiments of the present invention.

[0049] Figure 6 is the accuracy change curve of each method on the CIFAR-10 dataset in the embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention will be further described in detail below in conjunction with test examples and specific embodiments. However, this should not be understood as limiting the scope of the present invention to the following embodiments, and all technologies implemented based on the present invention fall within the scope of the present invention.

[0051] Example 1

[0052] The lightweight federated learning aggregation method based on genetic algorithms proposed in this paper can produce relatively good application effects in IoT neighborhoods with extreme data and the current existence of some high-quality public data sets. The genetic algorithm (GenFed method) can optimize model training through a distributed learning mechanism while ensuring communication efficiency and device-side computing power. It is suitable for large-scale IoT application scenarios such as smart homes, smart cities, and smart manufacturing. The performance verification of a specific lightweight federated learning aggregation method based on genetic algorithms is as follows.

[0053] To evaluate this plug-and-play and easily scalable federated learning design, this paper integrates the GenFed method into existing federated algorithms to explore the practical effectiveness of the genetic mechanism in the federated framework. Specifically, this paper selects the classic federated method FedAvg[3], the federated training acceleration method FedUmf

[30] , and the latest federated learning method FedAU

[18] based on weight design as comparison schemes to explore the performance of these methods in different task scenarios before and after integrating the GenFed mechanism. The specific steps are as follows:

[0054] S1. The server randomly selects a set amount of IoT device data from the global verification set to supplement the data of the client that sent the data supplement request;

[0055] S2. The server implements the client selection function through the communication interface module and distributes the current global model of the Internet of Things device to the client.

[0056] S3, the server receives the global model returned after training by the client through the communication interface module

[0057] S4, the server side selects the global model returned by the model selection module Fitness scoring is performed based on the performance in the global validation set;

[0058] S5: The server selects the top ρ in performance ranking based on the fitness score using a binary heap data structure. t The returned global model

[0059] S6. The server aggregates the selected returned global models to form a new global model through the model aggregation module.

[0060] The lightweight genetic algorithm-based federated learning aggregation method further requires looping steps S2 to S7, and ends the loop when the global model reaches a set accuracy or the total number of communication rounds is exhausted.

[0061] In step S1, the server needs to wait for a fixed set time t after receiving the data supplement request sent by the client. wait After completing the data transmission, the server needs to wait for a fixed set time t wait , reserving time for the client to receive data.

[0062] From the perspective of the client, first, non-malicious clients will actively participate in data supplementation. Clients with limited data can significantly improve the performance of the model obtained by local training by using the data provided by the server. In contrast, for clients with a large amount of data, the contribution of server-side data to the enhancement of model performance is negligible, and may even lead to an increase in the demand for the client's own local computing resources. Therefore, for these clients, deceiving the server in this process will have a negative impact on their own training. Secondly, malicious clients may regard the data distribution on the server as an opportunity to launch an attack and use the data obtained at this stage to carry out attacks. However, the process of distributing data on the server side is to randomly select a part of the data and send it to the client. Therefore, for any client i,j that issues a data support request, the supplementary data they receive and global data The following relationship is satisfied:

[0063]

[0064] This means that malicious clients not only cannot obtain complete global verification data, but also cannot obtain any data that is identical to that obtained by other clients. This makes their attack behavior at this stage somewhat difficult. Furthermore, the initialization phase only occurs once and accounts for a very small proportion of the framework of this invention, so the attack and defense behavior at this stage are not the focus of this invention.

[0065] The global validation set must meet the following key requirements: (1) the server must be able to fairly characterize the client's environment; (2) the server needs to evaluate the model returned by the client based on this characterization and select the models to participate in the aggregation accordingly. Based on these requirements, the present invention introduces a public dataset to be used as a global validation set. This dataset has the characteristics of balanced distribution and privacy independence, and provides a standardized model evaluation benchmark for the client during the training phase. The server uses this dataset to evaluate the models returned by the client in each round of training, and selects the models with excellent performance to participate in the aggregation, thereby constructing a new global model.

[0066] The scores obtained in step S5 are ranked in the top t The global model is added to the new set middle, and The following relationship exists:

[0067]

[0068] Where: Represents the set of clients selected to participate in this communication round.

[0069] The new global model in step S6 uses the new set Perform aggregation as follows:

[0070]

[0071] Where: represents the new global model; The global model returned by the server according to the client; represents the new set; p i represents the aggregate weight of the model from client i.

[0072] 8. A lightweight federated learning aggregation method based on genetic algorithm according to claim 1, characterized in that the number of the global models participating in the aggregation in step S6 is p t , ρ t It can be set according to different situations, as follows:

[0073] The first setting method is to maintain the number of aggregations at a constant value, denoted as:

[0074] ρ t =ρ max

[0075] Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregates during the experiment;

[0076] The second setting method uses a power function, and the aggregate quantity change process can be expressed as:

[0077] ρ t =ρ max *(1-b t )+1while b∈(0,1)

[0078] Where: ρ tis the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregations during the experiment; b is the base number that needs to be set before the experiment;

[0079] The third setting method is based on a linearly changing number of aggregations, as follows:

[0080]

[0081] Where: ρ t is the number of the global models participating in the aggregation; ρ max represents the maximum value of the number of aggregations during the experiment; c represents the maximum number of iterations of the maximum change in the number of aggregations;

[0082] The fourth design method is based on the sinusoidal variation of the aggregate quantity. The first sinusoidal variation of the aggregate quantity model follows a period of The sine function is written as:

[0083]

[0084] Another model uses a sine function with a complete π cycle, which is written as:

[0085]

[0086] Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregates during the experiment.

[0087] To evaluate the performance of the model, the present invention uses three datasets widely used in the field of federated learning: MNIST

[31] , Fashion-MNIST

[32] and CIFAR-10

[23] for experiments. Unless otherwise specified, the Dirichlet distribution

[33] with parameter is used in the experiment to simulate the data heterogeneity of 100 clients, and the CIFAR-10 dataset is used for performance evaluation. Since the research focus is on the design of the federated framework, the present invention only uses a simple convolutional neural network (CNN model) for training. For the experiments on the MNIST and Fashion-MNIST datasets, the present invention performed 2000 training rounds; in contrast, due to the greater difficulty of training the CIFAR-10 dataset, 4000 training rounds were performed. In addition, in each communication round, 10 clients are selected to participate in the training, and each client performs 5 rounds of local training.

[0088] To verify the effect of GenFed in improving training efficiency and model performance in different tasks, the present invention conducted experiments on multiple datasets. The results are as follows: Figure 2As shown in the figure, after the introduction of the GenFed mechanism, the performance of several traditional methods has been improved to varying degrees. In addition, the application of the GenFed mechanism significantly reduces the number of communication rounds required to achieve high performance. Specifically, on the MNIST dataset, GenFed only requires 1 / 15 of the FedAvg training cycles to achieve 97% accuracy; on the FashionMNIST dataset, GenFed only requires 1 / 35 of the FedAvg training cycles to achieve 80% accuracy; and on the CIFAR-10 dataset, GenFed only requires 1 / 10 of the FedAvg training rounds to achieve 45% accuracy. These results fully demonstrate the actual effect of GenFed in significantly improving model performance while accelerating the training process.

[0089] At the same time, in order to verify the effectiveness of the GenFed method under different client numbers, the present invention conducted the following Figure 3 The experiments shown in Figure 2 demonstrate that the GenFed approach consistently outperforms traditional methods in scenarios with a wide range of client counts. Of particular note, GenFed maintains stable and high performance when the number of clients exceeds 140. In contrast, the performance of traditional methods gradually deteriorates with increasing client counts, with accuracy dropping significantly. These experimental results demonstrate the GenFed approach's exceptional robustness and effectiveness in environments with a large number of clients, addressing the performance limitations of traditional methods in large client populations.

[0090] In order to verify the effect of different Non-IID levels on the performance of the GenFed method, the present invention conducted experiments based on different Dirichlet distribution parameters. Figure 4 As shown in the figure, the GenFed approach outperforms traditional approaches in most cases. In particular, under extreme conditions, due to the highly uneven data distribution across clients, traditional approaches struggle to maintain high accuracy across all clients. However, the proposed approach improves average client accuracy by 11% compared to traditional approaches under these conditions, demonstrating GenFed's robustness in dealing with severe data imbalance.

[0091] To evaluate the effectiveness of the GenFed method in the presence of malicious clients, we conducted an attack experiment based on the CIFAR-10 dataset. The specific malicious client implementation is as follows:

[0092] 1) Label Flipping (LF)

[34] : The malicious client modifies the labels of the local training data and flips the labels from 0 to 9, that is, the label x becomes 9-x.

[0093] 2) Inner Product Manipulation Attack (IPM)

[35] : A malicious client submits a negative vector of the normal client’s average gradient to the server, thereby causing destructive effects on the model update process.

[0094] 3) Mimic Attack

[36] : A malicious client submits a randomly selected copy of a normal client’s gradient to the server to mimic the behavior of a benign client.

[0095] The experimental results are as follows Figure 5 As shown in Figure 3, in three different attack scenarios, the performance degradation of the GenFed method is significantly lower than that of the FedAvg method. This shows that the integration of the GenFed mechanism effectively enhances the robustness of the federated learning algorithm in dealing with malicious client attacks.

[0096] Figure 2 The results show that the GenFed method achieves higher accuracy with fewer training rounds and converges significantly faster than traditional algorithms. To further analyze the impact of different parameter settings on the federated learning training process, Figure 6 The accuracy trends of various methods on the CIFAR-10 dataset during training are shown in detail. Among the different specific implementation methods, the GenFed models with dynamically changing parameters (including GenFed(2), GenFed(3), GenFed(4), and GenFed(5)) outperform the fixed GenFed(1) method in accelerating the training process and improving the final model accuracy. However, it should be noted that GenFed(5) exhibits significant instability in the later stages of training, affecting its overall reliability.

[0097] In view of the above findings, it is recommended to give priority to the GenFed(3) and GenFed(4) methods. These two methods show an excellent balance between fast convergence and final model performance, and have both high efficiency and excellent results, thus having wide applicability in various application scenarios.

[0098] The expected practical effects of deploying the GenFed approach in the above process are as follows:

[0099] 1. Significantly improves model convergence speed: By introducing a genetic mechanism, GenFed accelerates model training convergence, shortening training time compared to traditional federated learning methods in environments with highly heterogeneous data. For example, in the case of large IoT datasets, the number of communication rounds required for training is reduced, significantly improving model deployment efficiency.

[0100] 2. Enhanced accuracy and model performance: GenFed optimizes the model aggregation strategy to achieve higher prediction accuracy on different types of datasets.

[0101] 3. Improved robustness against malicious attacks: In federated learning, some participants may submit malicious updates, impacting the performance of the global model. GenFed effectively suppresses the negative impact of malicious participants on the global model through adaptive weight adjustment in a genetic mechanism, improving the system's robustness against attacks. It is suitable for scenarios facing security risks, such as the Internet of Things.

[0102] 4. Support for scalability in large-scale client environments: As the number of participating clients increases, the performance of traditional federated learning methods decreases significantly. However, GenFed, through the rational scheduling and balancing of resources, only slightly decreases its performance even with large-scale client participation. This enables GenFed to maintain efficient and stable operation in large-scale distributed systems of the Internet of Things.

[0103] In summary, the GenFed framework has broad application, particularly in distributed systems like the Internet of Things, which involve multiple clients, highly heterogeneous data, and limited resources. Its efficiency and robustness provide strong technical support for large-scale real-world scenarios.

Claims

1. A lightweight federated learning aggregation method based on genetic algorithm, characterized by: The system includes a server side, which specifically includes a communication interface module, a model selection module based on a genetic algorithm, and a model aggregation module. The server side implements the aggregation of the global model of the Internet of Things by the following steps: S1. The server randomly selects a set amount of IoT device data from the global verification set to supplement the data of the client that sent the data supplement request; S2. The server implements the client selection function through the communication interface module and distributes the current global model of the Internet of Things device to the client. S3, the server receives the global model returned after training by the client through the communication interface module S4, the server side selects the global model returned by the model selection module Fitness scoring is performed based on the performance in the global validation set; S5: The server selects the top ρ in performance ranking based on the fitness score using a binary heap data structure. t The returned global model S6. The server aggregates the selected returned global models to form a new global model through the model aggregation module. The number of the global models participating in the aggregation in step S6 is ρ t ,ρ t Set it according to different situations, as follows: The first setting method is to maintain the number of aggregations at a constant value, which is recorded as: r t =ρ max Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregates during the experiment; The second setting method uses a power function, and the aggregate quantity change process can be expressed as: r t =ρ max *(1-b t )+1while b∈(0,1) Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregations during the experiment; b is the base number that needs to be set before the experiment; The third setting method is based on a linearly changing number of aggregations, as follows: Where: ρ t is the number of the global models participating in the aggregation; ρ max represents the maximum value of the number of aggregations during the experiment; c represents the maximum number of iterations of the maximum change in the number of aggregations; The fourth design method is based on the sinusoidal variation of the aggregate quantity. The first sinusoidal variation of the aggregate quantity model follows a period of The sine function is written as: Another model uses a sine function with a complete π cycle, which is written as: Where: ρ t is the number of the global models participating in the aggregation; ρ max Indicates the maximum number of aggregates during the experiment.

2. A lightweight federated learning aggregation method based on genetic algorithm according to claim 1, characterized in that: The lightweight genetic algorithm-based federated learning aggregation method further requires looping steps S2 to S7, and ends the loop when the global model reaches a set accuracy or the total number of communication rounds is exhausted.

3. A lightweight federated learning aggregation method based on genetic algorithm according to claim 1, characterized in that: In step S1, the server needs to wait for a fixed set time t after receiving the data supplement request sent by the client. wait After completing the data transmission, the server needs to wait for a fixed set time t wait .

4. A lightweight federated learning aggregation method based on genetic algorithm according to claim 1, characterized in that: The server randomly selects a set amount of IoT device data from the global verification set to supplement data for the client that sends the data supplement request, which only happens once.

5. The lightweight genetic algorithm-based federated learning aggregation method according to claim 1, characterized in that: The global validation set is an introduced public data set, which provides a model evaluation benchmark for the client training in step S4; the global validation set also serves as a benchmark for fitness scoring in step S5.

6. A lightweight genetic algorithm-based federated learning aggregation method according to claim 1, characterized in that: The scores obtained in step S5 are ranked in the top t The global model is added to the new set middle, and The following relationship exists: Where: Represents the set of clients selected to participate in this communication round.

7. A lightweight federated learning aggregation method based on genetic algorithm according to claim 1, characterized in that: The new global model in step S6 uses the new set Perform aggregation as follows: Where: represents the new global model; The global model returned by the server according to the client; represents the new set; p i represents the aggregate weight of the model from client i.

Images