Regional communication local area network construction method based on digital twinning

Abstract

The invention relates to the technical field of regional communication local area networks, and discloses a regional communication local area network construction method based on digital twinning, comprising the following steps: step S101, collecting equipment information, performance data and fault records of equipment; step S102, preprocessing the equipment information, the performance data and the fault record of the equipment; step S103, constructing topological graph structure data; step S104, constructing a local area network scoring model according to the topological graph structure data; and step S105, generating a configuration scheme of the local area network in the area G by combining an improved crown porcupine optimization algorithm with the local area network scoring model. According to the method, the score of the connection between the devices is obtained by constructing the local area network scoring model, and the local area network configuration with the lowest delay and the lowest packet loss rate is obtained in combination with the improved crown porcupine optimization algorithm, so that the dynamic simulation of the regional communication local area network is realized, the prediction of the network performance is realized, and the detection of potential faults is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of regional communication local area networks, and more specifically, to a method for constructing a regional communication local area network based on digital twins. Background Art

[0002] With the rapid development of science and technology, regional communication LANs have become increasingly important in modern society. Whether in corporate offices, educational institutions or urban infrastructure, LANs play a key role. With the popularization of technologies such as the Internet of Things, big data, and cloud computing, the surge in the number of network devices and data traffic has made efficient and stable LANs the core infrastructure to ensure smooth information transmission. People's demand for LANs has shifted from simple connections to a comprehensive pursuit of high bandwidth, low latency, high reliability, and flexible scalability.

[0003] At present, the construction of regional communication LAN mainly relies on traditional network design and deployment technologies. These technologies usually include: demand analysis, network topology design, equipment selection and configuration, physical backup line, network testing and optimization, etc. During the design process, network engineers will use professional network simulation software for simulation verification and use test tools to evaluate the performance and stability of the network. These methods rely on the rich experience of engineers and the relevant functions of professional tools, and can meet the construction needs of LAN to a certain extent.

[0004] However, as the complexity of the network environment increases, the shortcomings of this traditional method are gradually becoming apparent. First, the traditional LAN construction method is usually based on a static model, which makes it difficult to dynamically reflect the state changes of the network in actual operation and cannot respond to emergencies in the network in real time; second, in the face of large-scale and complex networks, relying on manual experience and traditional simulation tools may lead to differences between the design and the actual deployment effect, as well as unreasonable resource allocation; in addition, the traditional method has shortcomings in fault prediction and prevention, mainly relying on post-detection and reactive maintenance, which can easily lead to network interruptions and reduced service quality. Summary of the invention

[0005] The present invention provides a method for constructing a regional communication local area network based on digital twins, which solves the technical problems in the related technology of not being able to dynamically reflect the state changes of the network in actual operation, unreasonable resource allocation, and being unable to prevent equipment failures.

[0006] The present invention provides a method for constructing a regional communication local area network based on digital twins, comprising the following steps: Step S101, collecting device information, performance data and fault records of M devices in a region G within a preset time period T at a preset time interval t, and representing the device information, performance data and fault record of each device by a first feature sequence; The nth sequence unit of the first feature sequence of the mth device represents the device information, performance data and fault record of the mth device at the nth time point, where 1≤m≤M, 1≤n≤N, and N=T / t; Equipment includes: switches, routers, servers and terminal equipment; Device information includes: device type, CPU model, memory capacity, hard disk type, interface type, connection cable type, IP address and port number; Performance data includes: uplink traffic, downlink traffic, latency, throughput, and packet loss rate; Fault records include: fault type and fault time of equipment fault; Step S102, preprocessing the device information, performance data and fault records of the M devices, converting the first feature sequence of each device into a second feature sequence representation; Step S103, constructing topology structure data according to the second characteristic sequences of the M devices; The topological graph structure data includes: nodes, initial features of nodes and edges; Establish a mapping relationship between devices and nodes; The initial feature of the node is represented by the second feature sequence of the device with which the mapping relationship is established; Step S104, constructing a local area network scoring model according to the topology structure data; The input of the LAN scoring model is the topology structure data, and the output is the score of each edge of the topology structure; Step S105, generating a configuration plan for the local area network in region G by combining the improved crown porcupine optimization algorithm with the local area network scoring model; The configuration scheme of the LAN in region G represents the connection relationship between M devices.

[0007] Furthermore, T, t, G, M, N, m, and n are custom parameters.

[0008] Furthermore, the pre-processing step includes: Step S201, for missing data, use the average value of N sequence units to fill; Step S202, calculating the average values ​​of uplink traffic, downlink traffic, delay, throughput and packet loss rate within a preset time period, creating a time series feature, and splicing it after the data processed in the above steps; Step S203, calculating the variance of the delay within a preset time period, creating a delay fluctuation feature, and splicing it after the data processed by the above steps; Step S204, calculating the average time interval between two faults, creating a fault interval feature, and splicing it after the data processed in the above steps; Step S205, normalizing the data processed in the above steps by using a z-score normalization method.

[0009] Furthermore, the ways to construct edges include: When two devices are physically connected, an edge is built between them; When a connection is established between two devices through a software protocol, an edge is built between the two devices; When two devices have the same model, an edge is built between the two devices.

[0010] Further, the LAN scoring model includes: a first hidden layer, a second hidden layer and a classifier; The first hidden layer includes: N first units and an extractor; The nth first unit inputs the nth sequence unit of the second feature sequence of each node and outputs the first updated feature; The extractor is used to use the first updated feature output by the Nth first unit as the initial feature of the node; The second hidden layer inputs the topological graph structure data and outputs graph update features, where each row vector of the graph update features corresponds to the second update feature of a node; The second updated feature of each node is input into the classifier, and the classification space of the classifier represents the score of each edge of the topological graph structure in the region G.

[0011] Furthermore, the first unit is constructed based on a long short-term memory network unit.

[0012] Furthermore, the formula for the second hidden layer is: ; ; ; ; ; Where S represents the graph update feature output by the second hidden layer, represents the second updated feature of the mth node, represents the initial features of the mth node, Represents the first weight parameter corresponding to the mth node, Represents the element value of the i-th row and j-th column of the adjacency matrix. The size of the adjacency matrix is ​​M×M. The elements of the adjacency matrix represent the connection relationship between nodes. The element value is represented by 0 or 1. 0 means that there is no connection relationship between the nodes, otherwise there is a connection relationship. represents the weighted adjacency matrix, represents the normalized degree matrix, Represents the degree matrix. The size of the degree matrix is ​​M×M. The elements of the degree matrix represent the number of edge connections with the node. and They represent the delay weight adjacency matrix and the packet loss rate adjacency matrix respectively. α and β are random numbers between 0 and 1, representing the delay weight parameter and the packet loss rate weight parameter respectively. and denote the delay and packet loss rate between node i and node j respectively, ReLU represents the ReLU activation function.

[0013] Furthermore, the loss function of the LAN scoring model adopts the cross entropy loss function.

[0014] Furthermore, the steps of the improved crested porcupine optimization algorithm used in step S104 include: Step S301, generating an initialization population with Q number of individuals meeting the constraint conditions, where Q is a custom parameter; The constraints include: Physical distance constraint: The physical distance between devices is limited, so the distance between nodes needs to be less than a set threshold maxDist. The physical distance constraint can be expressed as: ; Interface type and quantity constraints: When connecting devices, you need to consider the restrictions on interface type and quantity. The interface type and quantity constraints can be expressed as: ; Connection type constraints: Different devices support different connection types; ; Device compatibility constraints: There are compatibility issues between different devices. Device compatibility constraints can be expressed as: ; Unilateral connection constraint: There can be at most one edge between any two devices. The unilateral connection constraint can be expressed as: ; in Indicates the device and equipment The physical distance between them, maxDist represents the maximum allowed physical distance, Indicates the device and equipment Whether there is a connection between them, if yes, it is 1, if not, it is 0. Indicates the device The number of ports already in use, Indicates the device The number of available ports, Indicates the device and equipment Supported connection types. C represents the set of all allowed connection types. Indicates the device and equipment Whether it is compatible, if it is compatible, it is 1, if it is not compatible, it is 0. Indicates the device and equipment The number of edges between The individuals of the initialized population are represented by matrix coding, which is expressed as: ; in They respectively indicate whether there is an edge between the first node and the first node to the Mth node. Respectively indicate whether there is an edge between the 1st node to the Mth node and the 1st node, Indicates whether there is an edge between the Mth node and the Mth node. The element value of the matrix encoding is represented by 0 or 1, 0 means no edge, and 1 means yes edge. Step S302, constructing an objective function by combining the local area network scoring model to calculate and obtain the fitness value of each individual in the initialized population; Step S303, generating a random number ranging from 0 to 1, and when the random number is less than a first threshold, performing global exploration, otherwise performing local development; The first threshold is a custom variable parameter, which decreases as the number of iterations increases. The formula for the first threshold is: ; in represents the first threshold when the population iteration number is k+1, represents the first threshold when the number of population iterations is k, and β represents a decreasing factor between 0 and 1; Step S3041, during global exploration, a random number ranging from 0 to 1 is generated, and when the random number is less than a second threshold, the first defense strategy is adopted, otherwise the second defense strategy is adopted; Step S3042: during local development, a random number ranging from 0 to 1 is generated. When the random number is less than the second threshold, the third defense strategy is adopted, otherwise the fourth defense strategy is adopted. The formula for the first defense strategy is: ; ; The formula for the second defense strategy is: ; The formula for the third defense strategy is: ; ; ; ; The formula for the fourth defense strategy is: ; in represents the position of the i-th individual when the number of iterations is k+1, represents the position of the i-th individual when the number of iterations is k, represents random numbers that follow a normal distribution, , , , and rand is a random number between 0 and 1, represents the global optimal solution when the number of iterations is k, represents the position of the predator when the number of iterations is k, r is a random integer between 1 and Q, represents a binary vector consisting entirely of 0s, then , the predator no longer moves towards the crested porcupine, when When contains 1, the second defense strategy is based on two random individuals and To determine whether the predator is approaching or moving away from the crested porcupine, r1, r2, and r3 are random integers between 1 and Q. represents the first vector parameter used to control the search direction, represents the defense factor when the number of iterations is k, It represents the odor diffusion factor when the number of iterations is k, where k is the current number of iterations. is the maximum number of iterations, represents the fitness of the i-th individual when the number of iterations is k, As the second parameter, avoid division by 0. is the speed convergence factor, exp represents the exponential function; Step S305, after the current iteration is completed, the number of populations is reduced according to the cyclic population reduction technique, and the first threshold is updated; The formula for the cyclic population reduction technique is: ; in represents the number of individuals in the updated population, represents the minimum value of the population, represents the population size when the number of iterations is k, Indicates the loop variable, which is a custom parameter. Represents the remainder operation; Step S306, when the current number of iterations is greater than or equal to the maximum number of iterations, the individual with the largest fitness in the initialized population is used as the configuration scheme, otherwise, the process returns to step S302, where the maximum number of iterations is a custom parameter.

[0015] Furthermore, the calculation formula of the objective function is: ; Where Fitness represents the fitness value of the individual. represents the score of the edge between the mth node and the nth node obtained by the LAN scoring model, Represents the element value of the mth row and nth column of the individual code.

[0016] Furthermore, the packet loss rate and delay are obtained by sending test data packets between the two devices, and the packet loss rate and delay are weightedly summed to obtain a first score, and then the first score is normalized to obtain a score between the nodes that establish a mapping relationship with the two devices. The weight parameters for the weighted sum of the packet loss rate and the delay are all custom parameters, and the sum of the weight parameters corresponding to the packet loss rate and the weight parameters corresponding to the delay is 1.

[0017] The beneficial effects of the present invention are as follows: the present invention obtains the score of the connection between devices by constructing a local area network scoring model, and then obtains the local area network configuration with the lowest delay and the lowest packet loss rate through an improved crown porcupine optimization algorithm, thereby realizing dynamic simulation of the regional communication local area network, realizing the prediction of network performance, and realizing the detection of potential faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flow chart of a method for constructing a regional communication local area network based on digital twins of the present invention; Figure 2 It is a flow chart of preprocessing the equipment information, performance data and fault records of the equipment according to the present invention; Figure 3 It is a schematic diagram of the method of constructing edges in the topological graph structure data of the present invention; Figure 4 It is a flow chart of the improved crested porcupine optimization algorithm of the present invention. DETAILED DESCRIPTION

[0019] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0020] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in one or more embodiments of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0021] like Figure 1 to Figure 4 As shown, a method for constructing a regional communication LAN based on digital twins includes the following steps: Step S101, collecting device information, performance data and fault records of M devices in a region G within a preset time period T at a preset time interval t, and representing the device information, performance data and fault record of each device by a first feature sequence; The nth sequence unit of the first characteristic sequence of the mth device represents the device information, performance data and fault record of the mth device at the nth time point, where 1≤m≤M, 1≤n≤N, and N=T / t; Equipment includes: switches, routers, servers and terminal equipment; Device information includes: device type, CPU model, memory capacity, hard disk type, interface type, connection cable type, IP address and port number; Performance data includes: uplink traffic, downlink traffic, latency, throughput, and packet loss rate; Fault records include: fault type and fault time of equipment fault; Step S102, preprocessing the device information, performance data and fault records of the M devices, converting the first feature sequence of each device into a second feature sequence representation; Step S103, constructing topology structure data according to the second characteristic sequences of the M devices; The topological graph structure data includes: nodes, initial features of nodes and edges; Establish a mapping relationship between devices and nodes; The initial feature of the node is represented by the second feature sequence of the device with which the mapping relationship is established; Step S104, constructing a local area network scoring model according to the topology structure data; The input of the LAN scoring model is the topology structure data, and the output is the score of each edge of the topology structure; Step S105, generating a configuration plan for the local area network in region G by combining the improved crown porcupine optimization algorithm with the local area network scoring model; The configuration scheme of the LAN in region G represents the connection relationship between M devices.

[0022] In one embodiment of the present invention, T, t, G, M, N, m and n are custom parameters. Preferably, T is set to 24 hours and t is set to 1 hour.

[0023] In one embodiment of the present invention, Figure 2 As shown, the preprocessing steps include: Step S201, for missing data, use the average value of N sequence units to fill; Step S202, calculating the average values ​​of uplink traffic, downlink traffic, delay, throughput and packet loss rate within a preset time period, creating a time series feature, and splicing it after the data processed in the above steps; Step S203, calculating the variance of the delay within a preset time period, creating a delay fluctuation feature, and splicing it after the data processed by the above steps; Step S204, calculating the average time interval between two faults, creating a fault interval feature, and splicing it after the data processed in the above steps; Step S205, normalizing the data processed in the above steps by using a z-score normalization method.

[0024] In one embodiment of the present invention, Figure 3 As shown, the ways to construct edges include: When two devices are physically connected, an edge is built between them; When a connection is established between two devices through a software protocol, an edge is built between the two devices; When two devices have the same model, an edge is built between the two devices; In one embodiment of the present invention, the LAN scoring model includes: a first hidden layer, a second hidden layer and a classifier; The first hidden layer includes: N first units and an extractor; The nth first unit inputs the nth sequence unit of the second feature sequence of each node and outputs the first updated feature; The extractor is used to use the first updated feature output by the Nth first unit as the initial feature of the node; The second hidden layer inputs the topological graph structure data and outputs graph update features, where each row vector of the graph update features corresponds to the second update feature of a node; The second updated feature of each node is input into the classifier, and the classification space of the classifier represents the score of each edge of the topological graph structure in the region G.

[0025] In one embodiment of the present invention, the first unit is constructed based on a long short-term memory network unit.

[0026] In one embodiment of the present invention, the formula of the second hidden layer is: ; ; ; ; ; Where S represents the graph update feature output by the second hidden layer, represents the second updated feature of the mth node, represents the initial features of the mth node, Represents the first weight parameter corresponding to the mth node, Represents the element value of the i-th row and j-th column of the adjacency matrix. The size of the adjacency matrix is ​​M×M. The elements of the adjacency matrix represent the connection relationship between nodes. The element value is represented by 0 or 1. 0 means that there is no connection relationship between the nodes, otherwise there is a connection relationship. represents the weighted adjacency matrix, represents the normalized degree matrix, Represents the degree matrix. The size of the degree matrix is ​​M×M. The elements of the degree matrix represent the number of edge connections with the node. and They represent the delay weight adjacency matrix and the packet loss rate adjacency matrix respectively. α and β are random numbers between 0 and 1, representing the delay weight parameter and the packet loss rate weight parameter respectively. and denote the delay and packet loss rate between node i and node j respectively, ReLU represents the ReLU activation function.

[0027] In one embodiment of the present invention, the loss function of the local area network scoring model adopts a cross entropy loss function.

[0028] In one embodiment of the present invention, training samples of the local area network scoring model are obtained by manual labeling.

[0029] In one embodiment of the present invention, Figure 4 As shown, the steps of the improved crested porcupine optimization algorithm used in step S104 include: Step S301, generating an initialization population with Q individuals meeting the constraint conditions, where Q is a custom parameter, preferably, Q is set to 20; The constraints include: Physical distance constraint: The physical distance between devices is limited, so the distance between nodes needs to be less than a set threshold maxDist. The physical distance constraint can be expressed as: ; Interface type and quantity constraints: When connecting devices, you need to consider the restrictions on interface type and quantity. The interface type and quantity constraints can be expressed as: ; Connection type constraints: Different devices support different connection types; ; Device compatibility constraints: There are compatibility issues between different devices. Device compatibility constraints can be expressed as: ; Unilateral connection constraint: There can be at most one edge between any two devices. The unilateral connection constraint can be expressed as: ; in Indicates the device and equipment The physical distance between them, maxDist represents the maximum allowed physical distance, Indicates the device and equipment Whether there is a connection between them, if yes, it is 1, if not, it is 0. Indicates the device The number of ports already in use, Indicates the device The number of available ports, Indicates the device and equipment Supported connection types. C represents the set of all allowed connection types. Indicates the device and equipment Whether it is compatible, if it is compatible, it is 1, if it is not compatible, it is 0. Indicates the device and equipment The number of edges between The individuals of the initialized population are represented by matrix coding, which is expressed as: ; in They respectively indicate whether there is an edge between the first node and the first node to the Mth node. Respectively indicate whether there is an edge between the 1st node to the Mth node and the 1st node, Indicates whether there is an edge between the Mth node and the Mth node. The element value of the matrix encoding is represented by 0 or 1, 0 means no edge, and 1 means yes edge. For example The value 1 indicates that there is an edge between the 5th node and the 3rd node; Step S302, constructing an objective function by combining the local area network scoring model to calculate and obtain the fitness value of each individual in the initialized population; Step S303, generating a random number ranging from 0 to 1, and when the random number is less than a first threshold, performing global exploration, otherwise performing local development; The first threshold is a custom variable parameter, which decreases as the number of iterations increases. The formula for the first threshold is: ; in represents the first threshold when the population iteration number is k+1, represents the first threshold when the number of population iterations is k, β represents a decreasing factor between 0 and 1, preferably, the decreasing factor is set to 0.99, and the first threshold is set to 0.8; Step S3041, during global exploration, a random number ranging from 0 to 1 is generated, and when the random number is less than a second threshold, the first defense strategy is adopted, otherwise the second defense strategy is adopted; Step S3042: during local development, a random number ranging from 0 to 1 is generated. When the random number is less than the second threshold, the third defense strategy is adopted, otherwise the fourth defense strategy is adopted. The formula for the first defense strategy is: ; ; The formula for the second defense strategy is: ; The formula for the third defense strategy is: ; ; ; ; The formula for the fourth defense strategy is: ; in represents the position of the i-th individual when the number of iterations is k+1, represents the position of the i-th individual when the number of iterations is k, represents random numbers that follow a normal distribution, , , , and rand is a random number between 0 and 1, represents the global optimal solution when the number of iterations is k, represents the position of the predator when the number of iterations is k, r is a random integer between 1 and Q, represents a binary vector consisting entirely of 0s, then , the predator no longer moves towards the crested porcupine, when When contains 1, the second defense strategy is based on two random individuals and To determine whether the predator is approaching or moving away from the crested porcupine, r1, r2, and r3 are random integers between 1 and Q. represents the first vector parameter used to control the search direction, represents the defense factor when the number of iterations is k, It represents the odor diffusion factor when the number of iterations is k, where k is the current number of iterations. is the maximum number of iterations, represents the fitness of the i-th individual when the number of iterations is k, As the second parameter, avoid division by 0. is the speed convergence factor, exp represents the exponential function; Step S305, after the current iteration is completed, the number of populations is reduced according to the cyclic population reduction technique, and the first threshold is updated; The formula for the cyclic population reduction technique is: ; in represents the number of individuals in the updated population, Represents the minimum value of the population, preferably, Set to 10, represents the population size when the number of iterations is k, Represents a loop variable, preferably, Set to 2, Represents the remainder operation; Step S306, when the current number of iterations is greater than or equal to the maximum number of iterations, the individual with the largest fitness in the initialized population is used as the configuration scheme, otherwise return to step S302, where the maximum number of iterations is a custom parameter, preferably, the maximum number of iterations is set to 200.

[0030] In one embodiment of the present invention, the calculation formula of the objective function is: ; Where Fitness represents the fitness value of the individual. represents the score of the edge between the mth node and the nth node obtained by the LAN scoring model, Represents the element value of the mth row and nth column of the individual code.

[0031] Furthermore, the packet loss rate and delay are obtained by sending test data packets between the two devices, and the packet loss rate and delay are weightedly summed to obtain a first score, and then the first score is normalized to obtain a score between the nodes that establish a mapping relationship with the two devices. The weight parameters for the weighted sum of the packet loss rate and the delay are all custom parameters, and the sum of the weight parameters corresponding to the packet loss rate and the weight parameters corresponding to the delay is 1.

[0032] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A method for constructing a regional communication local area network based on digital twins, characterized in that: The following steps are involved: Step S101, collecting device information, performance data and fault records of M devices in a region G within a preset time period T at a preset time interval t, and representing the device information, performance data and fault record of each device by a first feature sequence; The nth sequence unit of the first characteristic sequence of the mth device represents the device information, performance data and fault record of the mth device at the nth time point, where 1≤m≤M, 1≤n≤N, and N=T / t; Equipment includes: switches, routers, servers and terminal equipment; Device information includes: device type, CPU model, memory capacity, hard disk type, interface type, connection cable type, IP address and port number; Performance data includes: uplink traffic, downlink traffic, latency, throughput, and packet loss rate; Fault records include: fault type and fault time of equipment fault; Step S102, preprocessing the device information, performance data and fault records of the M devices, converting the first feature sequence of each device into a second feature sequence representation; Step S103, constructing topology structure data according to the second characteristic sequences of the M devices; The topological graph structure data includes: nodes, initial features of nodes and edges; Establish a mapping relationship between devices and nodes; The initial feature of the node is represented by the second feature sequence of the device with which the mapping relationship is established; Step S104, constructing a local area network scoring model according to the topology structure data; The input of the LAN scoring model is the topology structure data, and the output is the score of each edge of the topology structure; Step S105, generating a configuration plan for the local area network in region G by combining the improved crown porcupine optimization algorithm with the local area network scoring model; The configuration scheme of the LAN in region G represents the connection relationship between M devices.

2. According to a method for constructing a regional communication LAN based on digital twins according to claim 1, it is characterized in that: T, t, G, M, N, m, and n are custom parameters.

3. According to a method for constructing a regional communication LAN based on digital twins according to claim 1, it is characterized in that: The preprocessing steps include: Step S201, for missing data, use the average value of N sequence units to fill; Step S202, calculating the average values ​​of uplink traffic, downlink traffic, delay, throughput and packet loss rate within a preset time period, creating a time series feature, and splicing it after the data processed in the above steps; Step S203, calculating the variance of the delay within a preset time period, creating a delay fluctuation feature, and splicing it after the data processed by the above steps; Step S204, calculating the average time interval between two faults, creating a fault interval feature, and splicing it after the data processed in the above steps; Step S205, normalizing the data processed in the above steps by using a z-score normalization method.

4. According to a method for constructing a regional communication LAN based on digital twins according to claim 1, it is characterized in that: Ways to construct edges include: When two devices are physically connected, an edge is built between them; When a connection is established between two devices through a software protocol, an edge is built between the two devices; When two devices have the same model, an edge is built between the two devices.

5. According to a method for constructing a regional communication LAN based on digital twins according to claim 1, it is characterized in that: The LAN scoring model includes: a first hidden layer, a second hidden layer and a classifier; The first hidden layer includes: N first units and an extractor; The nth first unit inputs the nth sequence unit of the second feature sequence of each node and outputs the first updated feature; The extractor is used to use the first updated feature output by the Nth first unit as the initial feature of the node; The second hidden layer inputs the topological graph structure data and outputs graph update features, where each row vector of the graph update features corresponds to the second update feature of a node; The second updated feature of each node is input into the classifier, and the classification space of the classifier represents the score of each edge of the topological graph structure in the region G.

6. A method for constructing a regional communication LAN based on digital twins according to claim 5, characterized in that: The first unit is built on the basis of the long short-term memory network unit.

7. A method for constructing a regional communication LAN based on digital twins according to claim 5, characterized in that: The formula for the second hidden layer is: ； ； ； ； ； Where S represents the graph update feature output by the second hidden layer, represents the second updated feature of the mth node, represents the initial features of the mth node, Represents the first weight parameter corresponding to the mth node, Represents the element value of the i-th row and j-th column of the adjacency matrix. The size of the adjacency matrix is ​​M×M. The elements of the adjacency matrix represent the connection relationship between nodes. The element value is represented by 0 or 1. 0 means that there is no connection relationship between the nodes, otherwise there is a connection relationship. represents the weighted adjacency matrix, represents the normalized degree matrix, Represents the degree matrix. The size of the degree matrix is ​​M×M. The elements of the degree matrix represent the number of edge connections with the node. and They represent the delay weight adjacency matrix and the packet loss rate adjacency matrix respectively. α and β are random numbers between 0 and 1, representing the delay weight parameter and the packet loss rate weight parameter respectively. and denote the delay and packet loss rate between node i and node j respectively, ReLU represents the ReLU activation function.

8. The method for constructing a regional communication LAN based on digital twin according to claim 1, characterized in that: The steps of the improved crested porcupine optimization algorithm used in step S104 include: Step S301, generating an initialization population with Q individuals meeting the constraint conditions, where Q is a custom parameter; The constraints include: Physical distance constraint: The physical distance between devices is limited, so the distance between nodes needs to be less than a set threshold maxDist. The physical distance constraint can be expressed as: ； Interface type and quantity constraints: When connecting devices, you need to consider the restrictions on interface type and quantity. The interface type and quantity constraints can be expressed as: ； Connection type constraints: Different devices support different connection types; ； Device compatibility constraints: There are compatibility issues between different devices. Device compatibility constraints can be expressed as: ； Unilateral connection constraint: There can be at most one edge between any two devices. The unilateral connection constraint can be expressed as: ； in Indicates the device and equipment The physical distance between them, maxDist represents the maximum allowed physical distance, Indicates the device and equipment Whether there is a connection between them, if yes, it is 1, if not, it is 0. Indicates the device The number of ports already in use, Indicates the device The number of available ports, Indicates the device and equipment Supported connection types. C represents the set of all allowed connection types. Indicates the device and equipment Whether it is compatible, if it is compatible, it is 1, if it is not compatible, it is 0. Indicates the device and equipment The number of edges between The individuals of the initialized population are represented by matrix coding, which is expressed as: ； in They respectively indicate whether there is an edge between the first node and the first node to the Mth node. Respectively indicate whether there is an edge between the 1st node to the Mth node and the 1st node, Indicates whether there is an edge between the Mth node and the Mth node. The element value of the matrix encoding is represented by 0 or 1, 0 means no edge, and 1 means yes edge. Step S302, constructing an objective function by combining the local area network scoring model to calculate and obtain the fitness value of each individual in the initialized population; Step S303, generating a random number ranging from 0 to 1, and when the random number is less than a first threshold, performing global exploration, otherwise performing local development; The first threshold is a custom variable parameter, which decreases as the number of iterations increases. The formula for the first threshold is: ； in represents the first threshold when the population iteration number is k+1, represents the first threshold when the number of population iterations is k, and β represents a decreasing factor between 0 and 1; Step S3041, during global exploration, a random number ranging from 0 to 1 is generated, and when the random number is less than a second threshold, the first defense strategy is adopted, otherwise the second defense strategy is adopted; Step S3042: during local development, a random number ranging from 0 to 1 is generated. When the random number is less than the second threshold, the third defense strategy is adopted, otherwise the fourth defense strategy is adopted. Step S305, after the current iteration is completed, the number of populations is reduced according to the cyclic population reduction technique, and the first threshold is updated; The formula for the cyclic population reduction technique is: ； in represents the number of individuals in the updated population, represents the minimum value of the population, represents the population size when the number of iterations is k, Indicates the loop variable, which is a custom parameter. Represents the remainder operation; Step S306, when the current number of iterations is greater than or equal to the maximum number of iterations, the individual with the largest fitness in the initialized population is used as the configuration scheme, otherwise, the process returns to step S302, where the maximum number of iterations is a custom parameter.

9. A method for constructing a regional communication LAN based on digital twins according to claim 8, characterized in that: The calculation formula of the objective function is: ； Where Fitness represents the fitness value of the individual. represents the score of the edge between the mth node and the nth node obtained by the LAN scoring model, Represents the element value of the mth row and nth column of the individual code.

10. A method for constructing a regional communication LAN based on digital twins according to claim 1, characterized in that: The packet loss rate and delay are obtained by sending test data packets between the two devices, and the packet loss rate and delay are weightedly summed to obtain a first score, and then the first score is normalized to obtain a score between the nodes that establish a mapping relationship with the two devices. The weight parameters for the weighted sum of the packet loss rate and the delay are all custom parameters, and the sum of the weight parameter corresponding to the packet loss rate and the weight parameter corresponding to the delay is 1.

Images