A method for determining the Internet of Things network based on an optimization strategy

Abstract

The present application relates to the field of Internet of Things technology. The present application provides an Internet of Things network determination method with an optimization strategy, characterized in that it includes: S10: defining the search space and evaluation index of the current Internet of Things network; S20: initializing the network structure parameters of the current Internet of Things network; S30: iteratively searching in the search space and performing optimization evaluation; S40: judging whether convergence occurs, and if convergence occurs, outputting the network structure parameters corresponding to the optimization evaluation result as the optimal network structure parameters of the current Internet of Things network. The present application can more comprehensively evaluate network performance and determine the optimal Internet of Things network by comprehensively considering various evaluation indicators.

Description

Technical Field

[0001] The present invention belongs to the field of Internet of Things, and in particular relates to an Internet of Things network determination method for an optimization strategy. Background Art

[0002] An IoT network is a network that connects and communicates devices in the Internet of Things (IoT). IoT networks allow various physical devices (such as sensors, smart home devices, industrial machines, etc.) to communicate and exchange data through the Internet or dedicated communication networks. These devices can be connected to the Internet through wired or wireless connections and can achieve remote monitoring, control and data transmission. IoT networks usually involve various technologies, including sensor technology, network communication technology, cloud computing and big data analysis, etc., to achieve intelligent interconnection between devices.

[0003] However, for the optimization of IoT networks, existing technologies usually only focus on a few evaluation indicators, such as throughput or latency, while ignoring other important indicators. This results in the network optimizing certain indicators while sacrificing other indicators, making it impossible to achieve comprehensive optimization. Summary of the invention

[0004] In view of the above-mentioned defects of the prior art, the present invention proposes an IoT network determination method with an optimization strategy. The technical solution steps designed by the present invention include:

[0005] S10: Define the search space and evaluation metrics of the current IoT network;

[0006] S20: Initialize the network structure parameters of the current Internet of Things network;

[0007] S30: iteratively search in the search space and perform optimization evaluation;

[0008] S40: Determine whether convergence has occurred. If convergence has occurred, output the network structure parameters corresponding to the optimization evaluation results as the optimal network structure parameters of the current Internet of Things network.

[0009] Preferably, the search space in S10 includes the number of network layers, the number of nodes in each layer, the type of activation function and the connection mode.

[0010] Preferably, the evaluation indicators in S10 include throughput, network delay, energy consumption and bandwidth utilization.

[0011] Preferably, S30 includes: using a global optimization algorithm to generate candidate solutions according to specified evaluation indicators and search space constraints, finding the optimal solution by gradually adjusting and iteratively optimizing the candidate solutions, and adjusting and updating the optimal solution according to its performance under the evaluation indicators.

[0012] Preferably, the optimization evaluation formula in S30 is as follows:

[0013]

[0014] In the formula, Ass is the evaluation index value, α i is the influence rate of the i-th search space, β l is the influence rate of the lth indicator, and S is the sample data set of each indicator in each search space.

[0015] Preferably, the α i The calculation formula is as follows:

[0016]

[0017] In the formula, γ i is the weight of the i-th search space, max(α i ) is the maximum influence rate in the search space.

[0018] Preferably, the β l The calculation formula is as follows:

[0019]

[0020] In the formula, δ l is the weight of the lth evaluation index, max(β l ) is the maximum impact rate among the evaluation indicators.

[0021] Preferably, before performing the optimization evaluation, it also includes performing a standardized conversion on each indicator sample data set in each search space, and the conversion formula is as follows:

[0022]

[0023]

[0024]

[0025] Where S1 is the sample data set of each indicator in each search space after standardized transformation, is the mean of S, σ is the standard deviation of S, n is the number of samples in S, and k is the sample index in S.

[0026] Beneficial effects:

[0027] 1. The present invention comprehensively considers various evaluation indicators. Compared with the method that only considers a single indicator, this solution comprehensively considers multiple indicators such as network delay, energy consumption, bandwidth utilization and throughput. This can more comprehensively evaluate network performance and ensure that the network reaches a better state in all aspects, rather than just pursuing performance improvement in one aspect;

[0028] 2. The present invention uses a dynamic search algorithm to search and optimize in a wider network structure space, and can find a network structure that is more suitable for actual needs, thereby improving network performance and efficiency;

[0029] 3. After the present invention introduces a dynamic optimization mechanism, the network can adjust parameters according to real-time environment and demand changes, thereby improving the flexibility and adaptability of the network and being able to better cope with different scenarios and workloads. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0031] The embodiments of the present invention are described in detail below. The following embodiments are implemented on the premise of the technical solution of the present invention, and detailed implementation methods and specific operation processes are given, but the protection scope of the present invention is not limited to the following embodiments.

[0032] The present invention designs a method for determining an Internet of Things network with an optimization strategy, and the technical solution comprises the following steps: Figure 1 As shown, specifically including:

[0033] S10: Define the search space and evaluation metrics of the current IoT network;

[0034] S20: Initialize the network structure parameters of the current IoT network;

[0035] S30: iteratively search in the search space and perform optimization evaluation;

[0036] S40: Determine whether convergence has occurred. If convergence has occurred, output the network structure parameters corresponding to the optimization evaluation results as the optimal network structure parameters of the current Internet of Things network.

[0037] Preferably, the search space in S10 includes the number of network layers, the number of nodes in each layer, the type of activation function and the connection mode.

[0038] Preferably, the evaluation indicators in S10 include throughput, network delay, energy consumption and bandwidth utilization.

[0039] Preferably, S30 includes: using a global optimization algorithm to generate candidate solutions according to specified evaluation indicators and search space constraints, finding the optimal solution by gradually adjusting and iteratively optimizing the candidate solutions, and adjusting and updating the optimal solution according to its performance under the evaluation indicators.

[0040] Preferably, the optimization evaluation formula in S30 is as follows:

[0041]

[0042] In the formula, Ass is the evaluation index value, α i is the influence rate of the i-th search space, β l is the influence rate of the lth indicator, and S is the sample data set of each indicator in each search space.

[0043] Specifically, each indicator sample data in each search space is generated according to a predefined network structure parameter combination.

[0044] Preferably, α i The calculation formula is as follows:

[0045]

[0046] In the formula, γ i is the weight of the i-th search space, max(α i ) is the maximum influence rate in the search space.

[0047] Preferably, β l The calculation formula is as follows:

[0048]

[0049] In the formula, δ l is the weight of the lth evaluation index, max(β l ) is the maximum impact rate among the evaluation indicators.

[0050] Specifically, the weights of the evaluation indicators (throughput, network latency, energy consumption, and bandwidth utilization) are calculated using the analytic hierarchy process (AHP). Each indicator is used as a different dimension in the judgment matrix of AHP, and then the relative importance between them is determined through the pre-defined network structure. Then, using the calculation method of AHP, the weight of each indicator relative to other indicators can be obtained.

[0051] Secondly, the selected search algorithm is used to iteratively search in the search space, and each network structure is evaluated and updated according to the evaluation index. Through continuous iterative optimization, the optimal solution can be gradually approached and the network performance can be improved. By calculating the influence rate of the index, the impact of different indicators on network performance can be reasonably weighed. This can more accurately reflect the performance of the network in various aspects, avoid the limitations that may exist in a single indicator, optimize the network performance, and evaluate the network performance more comprehensively. In addition, combined with the calculation of the influence rate of the search space, the weight of each search space is adaptively adjusted, so that the impact of different search spaces on network performance can be reasonably weighed, and the importance of each search space relative to other search spaces is obtained. The contribution of each search space to the overall evaluation index is clearly understood. Once the optimal solution is reached, the optimal network structure and optimal parameters corresponding to the search space can be obtained, which improves network performance evaluation and search efficiency.

[0052] Preferably, before performing the optimization evaluation, each indicator sample data set in each search space is also subjected to a standardized conversion, and the conversion formula is as follows:

[0053]

[0054]

[0055]

[0056] Where S1 is the sample data set of each indicator in each search space after standardized transformation, is the mean of S, σ is the standard deviation of S, n is the number of samples in S, and k is the sample index in S.

[0057] Specifically, since the difference between the sample data sets of each indicator in each search space and the numerical value of the evaluation indicator during the evaluation process is large, which will affect the subsequent convergence, the sample data sets of each indicator in each search space are standardized and transformed to ensure the accuracy and stability of the evaluation process.

[0058] The preferred specific embodiments of the present invention are described in detail above. It should be understood that ordinary technicians in the field can make many modifications and changes based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by technicians in the technical field based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for determining an Internet of Things network with an optimization strategy, characterized in that: include: S10: Define the search space and evaluation metrics of the current IoT network; S20: Initialize the network structure parameters of the current Internet of Things network; S30: iteratively search in the search space and perform optimization evaluation; S40: Determine whether convergence has occurred. If convergence has occurred, output the network structure parameters corresponding to the optimization evaluation result as the optimal network structure parameters of the current Internet of Things network. The S30 includes: using a global optimization algorithm to generate candidate solutions according to the specified evaluation index and search space constraints, finding the optimal solution by gradually adjusting and iteratively optimizing the candidate solutions, and adjusting and updating the optimal solution according to the performance of the optimal solution under the evaluation index; The optimization evaluation formula in S30 is as follows: In the formula, Ass is the evaluation index value, α i is the influence rate of the i-th search space, β l is the influence rate of the lth indicator, S is the sample data set of each indicator in each search space; The α i The calculation formula is as follows: In the formula, γ i is the weight of the i-th search space, max(α i ) is the maximum influence rate in the search space; The β l The calculation formula is as follows: In the formula, δ l is the weight of the lth evaluation index, max(β l ) is the maximum impact rate among the evaluation indicators.

2. The method for determining an Internet of Things network based on an optimization strategy according to claim 1, characterized in that: The search space in S10 includes the number of network layers, the number of nodes in each layer, the activation function type and the connection mode.

3. The method for determining an Internet of Things network based on an optimization strategy according to claim 1, characterized in that: The evaluation indicators in S10 include throughput, network delay, energy consumption and bandwidth utilization.

4. The method for determining an Internet of Things network based on an optimization strategy according to claim 1, characterized in that: Before the optimization evaluation is performed, each indicator sample data set in each search space is also subjected to standardized conversion. The conversion formula is as follows: Where S1 is the sample data set of each indicator in each search space after standardized transformation, is the mean of S, σ is the standard deviation of S, n is the number of samples in S, and k is the sample index in S.

Images