A wireless sensor routing method

Abstract

The present invention discloses a wireless sensor routing method, which first uses a clustering algorithm to calculate the optimal number of cluster heads and perform clustering, constructs a cluster structure, and then uses the constructed cluster structure to transmit data. The present invention solves the problems of uncertainty and uneven distribution of the number of cluster head nodes, reduces the corresponding energy consumption, and solves the problem of unnecessary energy consumption in the data transmission stage.

Description

Technical Field

[0001] The invention relates to the technical field of wireless sensors, in particular to a wireless sensor routing method. Background Art

[0002] A wireless sensor network is a network that integrates numerous sensor nodes with self-sensing capabilities through wireless communication technology. The nodes collect data from each other and send it to the base station, which then sends it to the network owner through satellite positioning technology. The units that make up a sensor node are: data collection unit, data transmission unit, data processing unit, and energy supply unit.

[0003] The data acquisition unit usually collects and converts information in the monitoring area, such as light intensity, atmospheric pressure and humidity; the data transmission unit mainly uses wireless communication and information exchange as well as sending and receiving the collected data information; the data processing unit usually handles the routing protocol and management tasks of all nodes and positioning devices; the energy supply unit will choose the form of micro batteries to reduce the area occupied by the sensor nodes. There are two types of nodes in the network: one is the aggregation node and the other is the sensor node. The aggregation node mainly collects and filters the data sent by the ordinary sensor nodes, and sends the data to the base station after fusion. The base station collects and processes the data. The sensor network realizes the three functions of data collection, processing and transmission. It is widely used in military, aviation, environment, medical, home, industry and other fields.

[0004] In the field of environmental monitoring, most sensor nodes are powered by batteries. Due to the complex environment, it is difficult to charge them in time. Therefore, how to improve the energy utilization of nodes is the key issue currently studied by scholars in this field. Existing scholars have proposed many methods for this problem. The earliest hierarchical routing protocol proposed is the LEACH algorithm proposed by Wendi Rabiner Heinzelman and other scholars. This algorithm first introduced the theoretical idea of ​​dynamic clustering and adopted a random cluster head election mechanism. Cluster heads are elected in each round of network operation. Ordinary member nodes choose appropriate clusters to join according to signal strength. However, during the election process, it is easy to ignore the remaining energy of nodes, which will lead to excessive network energy consumption during the election process and network paralysis caused by premature death of some nodes. In response to the energy consumption problem in the cluster head election process, relevant scholars proposed a LEACH-E energy-saving algorithm based on energy balance and clustering. In the initial stage, the algorithm takes into account the residual energy of the node and the average residual energy of the network. When selecting the cluster head, the node with a residual energy higher than the average energy of the network is selected as the cluster head. However, it does not solve the problem of uneven distribution of cluster heads and excessive energy consumption of single-hop transmission. There are also many improvements for heterogeneous sensor networks. In heterogeneous networks, someone proposed a clustering algorithm. In the cluster head selection stage, the residual energy and non-uniform competition mechanism are introduced, but the energy consumption problem in the transmission process is not considered.

[0005] The existing routing algorithms do not comprehensively consider the problem of residual energy causing nodes to die prematurely due to low energy, thus causing partial paralysis of the network, as well as the "hot zone" phenomenon close to the base station area. In addition, in different applications, there are different requirements for the election of cluster heads. A dynamic weighting factor should be introduced so that it can be dynamically adjusted. At the same time, the uncertainty of the number of cluster head nodes and the uneven distribution of cluster head nodes should be considered, which may lead to excessive energy consumption or data redundancy. At the same time, the problem of unnecessary energy consumption in the data transmission stage should be solved. Summary of the invention

[0006] To this end, the technical problem to be solved by the present invention is to provide a wireless sensor routing method, which solves the uncertainty and uneven distribution of the number of cluster head nodes, reduces the corresponding energy consumption, and solves the problem of unnecessary energy consumption in the data transmission stage.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] A wireless sensor routing method is based on a low-energy adaptive clustering hierarchical protocol for routing, specifically comprising the following steps:

[0009] S100) constructs a cluster structure, the specific operations are:

[0010] S110) Calculate the optimal number of cluster heads and divide the randomly distributed nodes in the wireless sensor network into a plurality of clusters according to the optimal number of cluster heads; wherein the optimal number of cluster heads is calculated by the following formula:

[0011]

[0012] Where N is the number of randomly distributed nodes in the wireless sensor network; M is the edge length of the wireless sensor network; To represent the energy consumption of the free space propagation model; is the energy consumption of the multipath fading propagation model; is the distance between the cluster head and the base station;

[0013] S120) In step S110), a node is elected as the cluster head in the cluster. When the cluster head election of all clusters is completed, the construction of the cluster structure is completed;

[0014] S200) Using the constructed cluster structure to transmit information between clusters.

[0015] In the wireless sensor routing method, in step S110), randomly distributed nodes in the wireless sensor network are divided into several clusters by the following steps:

[0016] S111) Determine the number of centroids according to the calculated optimal number of cluster heads, and select K nodes as cluster centers in the wireless sensor network, build a cluster with each cluster center, and then non-cluster center nodes select a suitable cluster to join according to the distance and calculate the mean of the cluster, where K is a non-zero natural number;

[0017] S112) calculating the distance between the non-cluster center nodes and the cluster center and regrouping the non-cluster center nodes according to the minimum distance formula;

[0018] S113) For each cluster, the average value of each feature of the cluster is obtained by a clustering algorithm, and a new cluster center in the cluster is selected again;

[0019] S114) Check whether the cluster center has converged. If the standard function difference converges to less than 0.01, the operation is terminated. Otherwise, return to step S112) and restart clustering until the standard function reaches convergence.

[0020] In the wireless sensor routing method described above, the minimum distance formula used in step S112) is:

[0021]

[0022] In the formula, is the center coordinate of the cluster object; are the coordinates of each non-cluster center node.

[0023] In the wireless sensor routing method, in step S120), after clustering is completed, a node is selected in each cluster as a cluster head. The specific operations are as follows:

[0024] The wireless sensor network base station sends the collected information to each node in each cluster structure area. After collecting the information, each node will run for cluster head. The election method is: add the current residual energy of the node and the average residual energy ratio of the wireless sensor network into the weight function formula, select the distance from the cluster center, and then select the cluster head. Less than 3R / 5 and remaining energy Greater than or equal to the remaining energy of the wireless sensor network 70% of the nodes are selected as cluster head candidates, and then the weight functions of the cluster head candidates are compared, and the cluster head candidates with smaller weight functions are comprehensively selected as cluster heads. R is the radius of each cluster structure area; the weight function value is calculated by the following weight function formula:

[0025]

[0026] In the formula, Indicates the distance from the node to the cluster center in the cluster; Represents the average distance from the node to the cluster center in the cluster; is the average residual energy of the wireless sensor network; is the current remaining energy of the node; It is a parameter used to control the value range, and its value range is (0,1).

[0027] In the wireless sensor routing method, in step S200), the ant colony algorithm is used to forward data when information is transmitted between clusters.

[0028] The wireless sensor routing method mentioned above incorporates the node residual energy and the residual energy average into the heuristic information when using the ant colony algorithm for data forwarding. middle.

[0029] The wireless sensor routing method described above uses the following transfer probability formula to select the next hop node when the ant colony algorithm is used for data forwarding:

[0030]

[0031] in, Calculated by the following formula:

[0032]

[0033] In the formula, represents the pheromone content on the edge (i, j); represents the heuristic information on the edge (i, j), The value is equal to the inverse of the distance between two nodes; represents a set of k next-hop nodes; α represents the pheromone weight; β represents the heuristic information weight; s represents a node among all the next-hop nodes; Represents the pheromone from the exploration node i to the next hop node; Represents the heuristic function from the exploration node to the next hop node; represents the Euclidean distance; represents the energy of exploring a node; represents the energy of node j; Indicates the energy consumption of sending data; Indicates the energy consumption of receiving data; represents the average energy consumption; φ represents the correction factor, and its value range is (0,1].

[0034] The wireless sensor routing method, when using the ant colony algorithm for data forwarding, updates pheromones based on the ant cycle model.

[0035] The wireless sensor routing method mentioned above integrates the energy, distance, average residual energy and initial energy of the node into the pheromone function when updating pheromone according to the ant cycle model, and defines the path with shorter path and higher average energy as the best path.

[0036] The wireless sensor routing method described above performs pheromone updates according to the ant cycle model by using the following formula:

[0037]

[0038]

[0039] In the formula, ρ represents the information volatility coefficient, and its value range is (0,1); represents the pheromone increment; r represents the number of cluster head election cycles; L b Indicates the distance between the current node and the next hop candidate node; represents the average remaining energy of the next-hop candidate node; E min represents the minimum energy of the next-hop candidate node; E0 represents the initial energy of the next-hop candidate node; Q is the pheromone concentration; μ is the correction coefficient, and its value range is (0,1).

[0040] The technical solution of the present invention achieves the following beneficial technical effects:

[0041] 1. Dynamically adjust the weighting factor according to the actual situation to obtain wireless sensor network data in uncertain situations such as the wild. Use the K-means algorithm to divide several cluster structures, and then elect cluster heads according to the proposed weight function. In the data transmission stage, the improved ant colony algorithm is used for data transmission, so that the intelligent optimization algorithm is combined with the clustering mechanism to establish multiple paths, and improve the transfer probability function and pheromone. By calculating and selecting an optimal path to send data to the base station, the energy consumption problem in the entire data establishment, acquisition, communication and transmission process is effectively optimized, and the speed of the wireless sensor network is greatly improved.

[0042] 2. The problem of excessive energy consumption in the data transmission stage caused by uniform distribution of cluster heads, diversity of data transmission paths between nodes within the cluster and cluster heads and aggregation nodes is solved. At the same time, the energy overhead problem caused by repeated election of cluster heads in the initial stage is solved. By selecting nodes with more residual energy and closer to the cluster center in each cluster as cluster heads, the distribution of cluster heads is more reasonable, and the position of cluster heads in the wireless sensor network is optimized, so that the wireless sensor network no longer consumes energy due to frequent selection of cluster heads, reducing unnecessary energy consumption of nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a composition diagram of a wireless sensor network;

[0044] Figure 2 It is a schematic diagram of the protocol stack of wireless sensor networks;

[0045] Figure 3 It is a schematic diagram of the topology of wireless sensor network;

[0046] Figure 4 Schematic diagram of a communication method based on LEACH;

[0047] Figure 5 It is a schematic diagram of the communication model in wireless sensor networks;

[0048] Figure 6 This is a flow chart of wireless sensor routing of the present invention. DETAILED DESCRIPTION

[0049] The present invention is further described below with reference to examples.

[0050] The wireless sensor network (WSN) involved in the present invention is composed of Figure 1As shown in the figure, it is a network composed of a large number of distributed nodes with low energy consumption. The network is mainly composed of sensor nodes, base stations, satellites and users. Sensor nodes collect data according to instructions, communicate data with each other within the communication radius, and finally send it to the base station in a single-hop or multi-hop communication mode. Compared with ordinary nodes, base stations have stronger information processing capabilities and are devices responsible for data collection. They send the collected data to users. After receiving the data, users transmit the data in the reverse direction and send information query and other task requests. The overall wireless sensor network is repeated so that terminal users can obtain corresponding data information, thereby better monitoring the regional information of the entire wireless sensor network and making the communication of the wireless sensor network go smoothly. The overall wireless sensor network architecture is relatively simple. Among them, the protocol stack is based on the reference OSI model and TCP / IP, combined with the characteristics of the WSN network itself, and is composed of the five-layer protocol in the traditional OSI model and the three platforms in the sensor network. Specifically, Figure 2 shown.

[0051] Since WSN is composed of a large number of randomly deployed nodes for data transmission, and the main function of the routing protocol is to find the communication path between the source node and the target node, due to the particularity of WSN, the node energy is limited. Therefore, in the design of WSN routing algorithm, the end point needs to pay attention to the energy consumption problem.

[0052] In the WSN of the present invention, a hierarchical routing protocol is used as the network routing protocol. Hierarchical routing is divided into cluster structures of different sizes by numerous random nodes. Through the division of the cluster structure, the protocol election process is carried out. The most used protocol is the LEACH protocol, and the present invention also optimizes and expands this protocol.

[0053] The Low Energy Adaptive Clustering Hierarchy (LEACH) protocol consists of cluster head nodes, common nodes, and base stations. In the wireless sensor network, cluster head election is first performed. Common nodes select appropriate clusters to join based on the perceived information. The wireless sensor network collects data and sends it to the cluster head node. The cluster head node summarizes the data and sends it to the base station. Its topology is as follows: Figure 3 As shown in Figure 2, LEACH divides the sensor network into several clusters in a “round” manner. Figure 4 The two stages complement each other and jointly complete the clustering work.

[0054] The communication model in wireless sensor networks is as follows: Figure 5As shown in Figure 1, the LEACH algorithm is used to select the corresponding mode according to the distance d between the sender and the receiver. If the distance d is less than the threshold d0, the free space mode is used to calculate the energy consumption. When the distance d is greater than the threshold d0, the multipath attenuation mode is selected to calculate the energy consumption. The energy consumption of bit data and the energy consumed for reception are:

[0055]

[0056]

[0057]

[0058] In the formula, E elec Indicates the energy required to send and receive each bit of data. is the length of the transmitted data, represents the energy consumption of the free space propagation model, Represents the energy consumption of the multipath fading propagation model.

[0059] In the present invention, Figure 6 As shown, the wireless sensor routing method is a routing method based on LEACH, which specifically includes the following steps:

[0060] S100) constructs a cluster structure, the specific operations are:

[0061] S110) Calculate the optimal number of cluster heads and divide the randomly distributed nodes in the wireless sensor network into a plurality of clusters according to the optimal number of cluster heads; wherein the optimal number of cluster heads is calculated by the following formula:

[0062]

[0063] Where N is the number of randomly distributed nodes in the wireless sensor network; M is the edge length of the wireless sensor network; To represent the energy consumption of the free space propagation model; is the energy consumption of the multipath fading propagation model; is the distance between the cluster head and the base station;

[0064] S120) In step S110), a node is elected as the cluster head in the cluster. When the cluster head election of all clusters is completed, the construction of the cluster structure is completed;

[0065] S200) Using the constructed cluster structure, an ant colony algorithm is used to transmit information between clusters.

[0066] The formula for calculating the optimal number of cluster heads is derived using a multipath attenuation propagation model. The reason is that the energy consumption of the cluster head node mainly comes from receiving data information collected by ordinary nodes in the cluster, fusing the information within the cluster, and sending the fused data information to the base station.

[0067] In the present invention, the algorithm used to calculate the optimal number of cluster heads in step S110) is the K-means algorithm. The K-means algorithm is a classic algorithm based on the partitioning method. Its main idea is to divide a given data set of X objects into k clusters, and each cluster obtained by partitioning is a cluster. In the wireless sensor network, k samples are randomly selected as cluster centers, and the remaining samples are added to the appropriate cluster according to the distance. After joining, the mean of the cluster is calculated, and then the category is reselected according to the similarity between each sample and the new cluster center, and it is continuously iterated until the criterion function converges to the minimum square error function value. The similarity evaluation index between the sample object and the cluster center is the distance between the two, and the Euclidean distance function is commonly used.

[0068] Specifically in the present invention, in step S110), randomly distributed nodes in the wireless sensor network are divided into several clusters through the following steps:

[0069] S111) Determine the number of centroids according to the calculated optimal number of cluster heads, and select K nodes as cluster centers in the wireless sensor network, build a cluster with each cluster center, and then non-cluster center nodes select a suitable cluster to join according to the distance and calculate the mean of the cluster, where K is a non-zero natural number;

[0070] S112) calculating the distance between the non-cluster center nodes and the cluster center and regrouping the non-cluster center nodes according to the minimum distance formula; wherein the minimum distance formula is:

[0071]

[0072] In the formula, is the center coordinate of the cluster object; is the coordinate of each non-cluster center node;

[0073] S113) For each cluster, the average value of each feature of the cluster is obtained by a clustering algorithm, and a new cluster center in the cluster is selected again;

[0074] S114) Check whether the cluster center has converged. If the standard function difference converges to less than 0.01, the operation is terminated. Otherwise, return to step S112) and restart clustering until the standard function reaches convergence.

[0075] In the K-means clustering algorithm, each cluster consists of multiple samples, and the location of the new cluster center of each cluster is calculated based on the eigenvalues ​​of all samples in the cluster. The specific steps are:

[0076] (1) Suppose there is a cluster containing m samples, each of which has n features (dimensions). For example, the features of a sample can be length, width, height, etc. Here, let the sample The characteristics of ,in, represents the number of sample M, Indicates the number of the feature.

[0077] (2) Calculate the average value of each feature in the cluster to determine the new cluster center position. For the jth feature, the new cluster center calculation formula is:

[0078]

[0079] Sum the values ​​of the jth feature of all m samples in the cluster, and then divide it by the number of samples m to get the value of the new cluster center in the jth dimension.

[0080] Using the clustering rules based on the K-means clustering algorithm to calculate the optimal number of cluster heads can make the cluster head distribution more reasonable and set the optimal number of clusters according to the different situations of different wireless sensor networks.

[0081] In step S120), after clustering is completed, a node is selected in each cluster as a cluster head. The specific operations are as follows:

[0082] The wireless sensor network base station sends the collected information to each node in each cluster structure area. After collecting the information, each node will run for cluster head. The election method is: add the current residual energy of the node and the average residual energy ratio of the wireless sensor network into the weight function formula, select the distance from the cluster center, and then select the cluster head. Less than 3R / 5 and remaining energy Greater than or equal to the remaining energy of the wireless sensor network 70% of the nodes are selected as cluster head candidates, and then the weight functions of the cluster head candidates are compared, and the cluster head candidates with smaller weight functions are comprehensively selected as cluster heads. R is the radius of each cluster structure area; the weight function value is calculated by the following weight function formula:

[0083]

[0084] In the formula, Indicates the distance from the node to the cluster center in the cluster; Represents the average distance from the node to the cluster center in the cluster; is the average residual energy of the wireless sensor network; is the current remaining energy of the node; It is a parameter used to control the value range, and its value range is (0,1).

[0085] Compared with the cluster head election method in the present invention, the traditional cluster head election does not take the residual energy of the cluster head node into consideration, which may cause nodes with too little residual energy to be elected as cluster head nodes, eventually leading to partial paralysis of the network and causing high energy consumption during the entire network transmission process.

[0086] After completing the construction of the cluster structure, the wireless sensor network enters the data transmission process, and the communication energy consumption in the sensor node energy consumption model accounts for the main part. In the process of data transmission, selecting an optimal path for data transmission can improve the performance of the network. In the present invention, the ant colony algorithm is introduced to forward data in the data transmission stage, and is combined with the LEACH algorithm in the WSN network to achieve the purpose of optimizing the routing path and improving network performance.

[0087] When information is transmitted between clusters, the commonly used method is multi-hop transmission between cluster heads. Therefore, the node's power consumption for transmitting data ETX and receiving data ERX becomes crucial. The original transfer probability formula in the ant colony algorithm only considers the influencing factor of the distance between nodes. However, in the routing protocol of WSN, the main performance indicator is energy consumption. Therefore, the present invention also incorporates the node residual energy and the residual energy average into the heuristic information. In the process, the probability of nodes with large residual energy being selected as the next hop node is increased. The improved transfer probability function will be selected based on the node spacing, the residual energy of the node, and the power consumption of the node sending and receiving data. At the same time, the importance of energy consumption in the data transmission stage is comprehensively considered, and a correction factor is introduced φ , and then adjust the correction factor according to different network environment requirements. φ Make dynamic adjustments to achieve dynamic control of the network.

[0088] Specifically, when the ant colony algorithm is used for data forwarding, the present invention incorporates the point residual energy and the residual energy average into the heuristic function, specifically, the following transition probability formula is used to select the next hop node:

[0089]

[0090] in, Calculated by the following formula:

[0091]

[0092] In the formula, represents the pheromone content on the edge (i, j); represents the heuristic information on the edge (i, j), The value is equal to the inverse of the distance between two nodes; represents the set of k next-hop nodes; α represents the pheromone weight; β represents the heuristic information weight; s represents all next-hop nodes; Represents the pheromone from the exploration node i to the next hop node; Represents the heuristic function from the exploration node to the next hop node; represents the Euclidean distance; represents the energy of exploring a node; represents the energy of node j; Indicates the energy consumption of sending data; Indicates the energy consumption of receiving data; represents the average energy consumption; φ Expressed as a correction factor, the value range is (0,1].

[0093] In the pheromone updating stage, there are usually three models in the ant colony algorithm that can be used for pheromone updating. The present invention adopts the ant cycle model for pheromone updating, and integrates the node's energy, distance, average residual energy, and initial energy into the pheromone function, and defines the path with a shorter path and a higher average energy as the optimal path.

[0094] Specifically, when updating pheromones according to the ant-week model, the pheromone update is performed by the following formula:

[0095]

[0096]

[0097] In the formula, ρ represents the information volatility coefficient, and its value range is (0,1); represents the pheromone increment; r represents the number of cluster head election cycles; L b Indicates the distance between the current node and the next hop candidate node; represents the average remaining energy of the next-hop candidate node; E min represents the minimum energy of the next-hop candidate node; E0 represents the initial energy of the next-hop candidate node; Q is the pheromone concentration; μ is the correction coefficient, and its value range is (0,1).

[0098] In the present invention, during the inter-cluster information transmission, the transfer probability function and the pheromone update function in the ant colony algorithm are optimized, and the sending and receiving energy consumption is introduced. By improving the pheromone update strategy, the pheromone update is determined by the path length and energy. By improving the probability selection and pheromone update rules, the optimal multi-hop path between the cluster head node and the sink node is searched, and the inter-cluster routing path is well planned, thereby alleviating the energy consumption problem in data transmission.

[0099] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the claims of this patent application.

Claims

1. A wireless sensor routing method, characterized in that: Routing is performed based on a low-energy adaptive clustering hierarchical protocol, specifically including the following steps: S100) constructs a cluster structure, the specific operations are: S110) calculating the optimal number of cluster heads and dividing the randomly distributed nodes in the wireless sensor network into a plurality of clusters according to the optimal number of cluster heads; The optimal number of cluster heads is calculated by the following formula: Where N is the number of randomly distributed nodes in the wireless sensor network; M is the edge length of the wireless sensor network; To represent the energy consumption of the free space propagation model; is the energy consumption of the multipath fading propagation model; is the distance between the cluster head and the base station; S120) In step S110), a node is selected as the cluster head in the cluster. When the cluster heads of all clusters are selected, the cluster structure is constructed. In step S120), after the clustering is completed, a node is selected as the cluster head in each cluster. The specific operations are: The wireless sensor network base station sends the collected information to each node in each cluster structure area. After collecting the information, each node will run for cluster head. The election method is: add the current residual energy of the node and the average residual energy ratio of the wireless sensor network into the weight function formula, select the distance from the cluster center, and then select the cluster head. Less than 3R / 5 and remaining energy Greater than or equal to 70% The nodes are selected as cluster head candidates, and then the weight functions of the cluster head candidates are compared, and the cluster head candidate nodes with smaller weight functions are selected as cluster heads. R is the radius of each cluster structure area; the weight function value is calculated by the following weight function formula: In the formula, Indicates the distance from the node to the cluster center in the cluster; Represents the average distance from the node to the cluster center in the cluster; is the average residual energy of the wireless sensor network; is the current remaining energy of the node; is a parameter used to control the value range, and its value range is (0,1); S200) Use the constructed cluster structure to transmit information between clusters, and use the ant colony algorithm to forward data when transmitting information between clusters; when using the ant colony algorithm to forward data, incorporate the node residual energy and the residual energy average into the heuristic information When the ant colony algorithm is used for data forwarding, the following transfer probability formula is used to select the next hop node: in, Calculated by the following formula: In the formula, represents the pheromone content on the edge (i, j); represents the heuristic information on the edge (i, j), The value is equal to the inverse of the distance between two nodes; represents a set of k next-hop nodes; α represents the pheromone weight; β represents the heuristic information weight; s represents a node among all the next-hop nodes; Represents the pheromone from the exploration node i to the next hop node; Represents the heuristic function from the exploration node to the next hop node; represents the Euclidean distance; represents the energy of exploring a node; represents the energy of node j; Indicates the energy consumption of sending data; Indicates the energy consumption of receiving data; represents the average energy consumption; φ represents the correction factor, and its value range is (0,1].

2. The wireless sensor routing method according to claim 1, characterized in that: In step S110), randomly distributed nodes in the wireless sensor network are divided into several clusters by the following steps: S111) Determine the number of centroids according to the calculated optimal number of cluster heads, and select K nodes as cluster centers in the wireless sensor network, build a cluster with each cluster center, and then non-cluster center nodes select a suitable cluster to join according to the distance and calculate the mean of the cluster, where K is a non-zero natural number; S112) calculating the distance between the non-cluster center nodes and the cluster center and regrouping the non-cluster center nodes according to the minimum distance formula; S113) For each cluster, the average value of each feature of the cluster is obtained by a clustering algorithm, and a new cluster center in the cluster is selected again; S114) Check whether the cluster center has converged. If the standard function difference converges to less than 0.01, the operation is terminated. Otherwise, return to step S112) and restart clustering until the standard function reaches convergence.

3. The wireless sensor routing method according to claim 2, characterized in that: The minimum distance formula used in step S112) is: In the formula, is the center coordinate of the cluster object; are the coordinates of each non-cluster center node.

4. The wireless sensor routing method according to claim 1, characterized in that: When the ant colony algorithm is used for data forwarding, pheromone is updated according to the ant cycle model.

5. The wireless sensor routing method according to claim 4, characterized in that: When updating pheromones according to the ant cycle model, the energy, distance, average residual energy and initial energy of the node are integrated into the pheromone function, and the path with shorter path and higher average energy is defined as the best path.

6. The wireless sensor routing method according to claim 5, characterized in that: When updating pheromones according to the ant cycle model, the pheromone is updated by the following formula: In the formula, ρ represents the information volatility coefficient, and its value range is (0,1); represents the pheromone increment; r represents the number of cluster head election cycles; L b Indicates the distance between the current node and the next hop candidate node; represents the average remaining energy of the next-hop candidate node; E min represents the minimum energy of the next-hop candidate node; E0 represents the initial energy of the next-hop candidate node; Q is the pheromone concentration; μ is the correction coefficient, and its value range is (0,1).

Images