WSN target coverage optimization method based on improved swarm intelligence algorithm

Abstract

The application belongs to the technical field of WSN and discloses a WSN target coverage optimization method based on an improved swarm intelligence algorithm. In view of the problems that the optimal scheme is obtained slowly, is prone to local optimum, and coverage redundancy exists, distribution is uneven, and thus resources are wasted and data quality and reliability are affected due to random deployment of the existing WSN node, the social spider algorithm containing an elite exclusion operator is used to solve the above problems. The method is used for constructing a coverage model, calculating the fitness of a spider individual after initialization, adjusting the position according to the position difference and the update parameter of the optimal spider, determining the elite individual and the exclusion degree during iteration, generating a new population through the exclusion operation, and ending the algorithm when the set iteration threshold is reached. The method improves the coverage efficiency and quality, avoids local optimum, converges fast and is stable in optimization, and provides an efficient scheme for WSN target coverage.

Description

TECHNICAL FIELD

[0001] The present application relates to a wireless sensor network (WSN) target coverage method, and proposes a method for solving the wireless sensor coverage problem by using an improved social spider algorithm, which can quickly find the best distribution scheme of the wireless sensor and show it. BACKGROUND

[0002] WSN is a network composed of a large number of interconnected wireless sensor nodes, which is used for monitoring and collecting environmental data. In the target coverage process of WSN, the deployment and layout of nodes are very critical, because they directly affect the performance and coverage range of the network. Therefore, target coverage is a very important problem in the field of WSN.

[0003] The social spider algorithm is a new swarm intelligence optimization algorithm that simulates the behavior of spiders in nature to divide labor, exchange information and reproduce offspring. The algorithm has the characteristics of simple structure, strong stability and easy understanding.

[0004] In many monitoring scenarios, such as environmental monitoring, industrial equipment operation state monitoring, security monitoring, etc., multiple sensor nodes are often deployed to collect data and monitor the state of multiple monitored targets. How to reasonably configure the monitoring relationship between sensor nodes and monitored targets, i.e., to determine which sensor nodes are responsible for monitoring which monitored targets, to achieve the highest monitoring coverage rate, to adapt the maximum fitness as the optimization goal, and to improve the adjustment operator to increase the population diversity and speed up the global optimization speed of the population, and to constantly update the population position, to cover all sensor nodes in the test area, to achieve coverage optimization, are the key problems to improve the performance and reliability of the monitoring system.

[0005] The coverage optimization problem of WSN is mainly described as randomly placing nodes in the specified monitoring area, and only covering the target area, without the need for full coverage of the standard area. There may be coverage redundancy and uneven distribution of nodes, which affects the quality and reliability of network monitoring and causes a lot of resource waste. Based on the traditional coverage method, the swarm intelligence algorithm is not used, and it is difficult to quickly and accurately find the optimal monitoring configuration scheme when facing complex and large-scale monitoring scenarios. Individuals are often prone to local optimization, resulting in insufficient coverage of some monitored targets or waste of sensor resources. When saving individual fitness values, the maximum fitness priority principle is adopted, and whether the individual with low fitness value continues to enter the cycle is not considered, which leads to a small coverage range.

[0006] Therefore, a WSN target coverage method with full coverage range and no coverage redundancy problem is needed.

[0007] Based on the above problems, this solution introduces elite and排挤 operatorsto optimize the search process of spiders. The elitesqueezing operator adjusts the search strategy and parameters of spiders according to the fitness value of spiders (i.e., coverage quality), enabling the algorithm to more effectively utilize the information of elite individuals during the search process, improving the convergence speed and coverage area of the algorithm. Summary of the Invention

[0008] The first technical solution is as follows:

[0009] A WSN target coverage optimization method based on an improved swarm intelligence algorithm, including:

[0010] Construct a WSN target coverage model and initialize it. Calculate the fitness values of each individual in the spider population of the improved social spider algorithm through a fitness function composed of the WSN target area coverage rate, and use the sensor position matrix corresponding to the individual with the largest fitness value as the optimal deployment plan for the wireless sensor network coverage optimization model.

[0011] According to the constructed WSN target coverage model, it includes:

[0012] Set the sensor node set F = {f1, f2,..., f i ,... f n}, where f i represents the i-th sensor node, where 0 < i < n. The monitored target set E = {e1, e2,..., e j ,... e m}, where e j represents the j-th monitored target, where 0 < j < m. Define the sensing range R s and communication range R c and R s < R c . Determine the coordinates of the sensor node f i in a two-dimensional coordinate system as and the position of the monitored target e j . Construct a coverage relationship matrix R between the sensor nodes and the monitored targets, where the element represents that the e j -th monitored target is not covered by the f i -th sensor node. If r e,f = 1, it means it is covered, that is, the distance is less than R s .

[0013]

[0014] Randomly initialize N spiders, that is, N groups of sensor positions;

[0015] Spider k O Position: Where 0 < O ≤ N;

[0016] Where each spider represents a sensor deployment scheme.

[0017] According to the described by the wireless sensor network target area coverage rate constitutes the fitness function to calculate the improved social spider algorithm of each individual of the spider population fitness value, comprising:

[0018] The selected target area size, the number of sensors and the sensor sensing radius are initialized, the correspondence between the target coverage concept and the improved social spider algorithm concept is determined, and the spider individual position is randomly initialized in the set monitoring area.

[0019] In the set monitoring area, the coverage rate of the sensor node is taken as the optimization target, the spider individual fitness value is obtained and sorted; wherein the individual spider k O The coverage matrix of As follows:

[0020]

[0021] Where the sensor node index is i, the sensing radius of the i-th sensor node is r i , the monitored target index is j; the total number of sensor nodes is n. Represents the i-th sensor node in the k O th scheme covers the j-th monitored target.

[0022] The two-dimensional coordinates of the i-th sensor node in the k O th spider individual are The two-dimensional coordinates of the j-th monitored target are

[0023] The fitness function of the individual spider k O The coverage rate is As follows, that is, covered by at least one sensor;

[0024]

[0025] According to the described individual corresponding to the sensor position matrix with the largest fitness value as the optimal deployment scheme of the wireless sensor network coverage optimization model, comprising:

[0026] The distance of the spider individual is calculated, and the position of the spider is updated; the vibration signal of the spider individual is propagated; the elite operator, the exclusion strategy and the position of the remaining spiders are reserved, and finally new spiders are randomly generated to join the group; it is judged whether the set improvement social spider algorithm iteration threshold is reached, if not, the above operations are cycled, otherwise, the optimal spider individual position and the optimal fitness are output, and the optimal deployment scheme of the sensor node is obtained.

[0027] According to the distance calculation and position update of the spider individual, the spider transmits information through vibration, the high fitness spider attracts other spiders through strong vibration, and the low fitness spider emits weak vibration and is repelled. It is assumed that the position coordinates of the target spider k t are The position coordinates of the other single spider k v are Wherein the total number of sensor nodes is n, wherein 0 t The two-dimensional coordinates of the i-th sensor node in the k t th spider individual are The x-axis coordinate of the i-th sensor node of the spider k t is The y-axis coordinate of the i-th sensor node of the spider k t is

[0028] The distance between two spiders is measured by considering the positions of all sensors As follows:

[0029]

[0030] The position update formula is:

[0031]

[0032] Wherein The new coordinates of the i-th sensor node of the target spider k t are The old coordinates of the i-th sensor node of the target spider k t are best The position of the i-th sensor corresponding to the current highest fitness spider is x t , α and β are weight coefficients for controlling the exploration and development of the spider, and rand() is a uniform random number generation function in the interval [0, 1].

[0033] Wherein the influence of other spiders on it is social influence:

[0034]

[0035] where V is the other spider k v the vibration signal intensity of the target spider k t the x-axis coordinate of the ith sensor node of the spider k v the x-axis coordinate of the ith sensor node of the spider k t

[0036] According to the vibration signal propagation of the spider individual, the vibration signal intensity formula is as follows:

[0037]

[0038] where, the fitness of the signal generating spider k v the distance between the spiders k v t

[0039] According to the elite, the exclusion strategy is adopted, and the position of the remaining spiders is reserved, and finally new spiders are randomly generated to join the group, including:

[0040] Select the spider individual with high fitness as the elite according to the elite ratio Q (0 < Q < 1), which is not excluded;

[0041] Let the positions of spiders k1 and k2 be:

[0042]

[0043] Then the similarity is:

[0044]

[0045] where C is the number of corresponding elements that are the same in the positions of the two spiders.

[0046] Calculate the similarity Sim(k1, k2) between individuals, where k1 and k2 are the position vectors of the two spiders, respectively;

[0047] Set the exclusion threshold: SimMAX (0 ≤ SimMAX ≤ 1);

[0048] Exclude the spider individual with the highest exclusion degree, and replace it with a new spider:

[0049]

[0050] ​​​​​​In the formula, PAI represents the exclusion operation, the spider with low fitness in the spider is excluded and replaced with a newly generated spider, and it is determined whether the set elite, the exclusion spider algorithm iteration threshold is reached, that is, (the maximum iteration number is reached, the coverage rate reaches the preset threshold, or the required solution is found), and the algorithm ends, and the optimal deployment scheme is output. If the algorithm threshold is not reached, the newly generated spider fitness is calculated, the spider attraction is re-substituted and calculated, and the above operation is sequentially and circularly performed.

[0051] The beneficial effects of the present application are as follows: the present application can effectively avoid falling into a local optimal solution by introducing an elite and exclusion operator, and enhances the search capability in a complex solution space, thereby improving the coverage efficiency and quality of the monitored target by the sensor node, and making the finally obtained sensor node deployment scheme closer to the global optimum. BRIEF DESCRIPTION OF DRAWINGS

[0052] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and serve to explain the principles of the application. In the drawings:

[0053] Figure 1 Flowchart of the WSN target coverage optimization method based on the improved swarm intelligence algorithm;

[0054] Figure 2 Simulation comparison diagram of the coverage range of the WSN target coverage method;

[0055] Figure 3 Abstract diagram of the WSN target coverage optimization method based on the improved swarm intelligence algorithm. DETAILED DESCRIPTION

[0056] In order to make the technical solutions and advantages in the embodiments of the present application clearer, the following further describes the exemplary embodiments of the present application with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and are not an exhaustive enumeration of all embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0057] Embodiment: Reference Figure 1 and Figure 2 The WSN target coverage optimization method based on the swarm intelligence algorithm in the present embodiment is described in detail, and specifically includes the following steps:

[0058] Step 1: Construct the WSN target coverage model and initialize, set the sensor node set F = {f1, f2, … f 15}, for example, f5 represents the 5th sensor node, the monitored target set E = {e1, e2, …, e100 For example, e5 represents the 5th monitored target, and the initial population size is 40, i.e., 40 sensor locations. Define the sensing range R of the sensor node. s =130m and communication range R c =15m, sensor node f is determined using a two-dimensional coordinate system. i The coordinates are With the monitored target e j Location And construct the coverage relationship matrix R between sensor nodes and monitored targets, where element r e,f ∈{0,1}, for example r 5,5 =0 means the 5th monitored target is not covered by the 5th sensor node, r 5,5 =1 indicates that it is covered, meaning the distance is less than R. s .

[0059]

[0060] Step 2: Within the designated monitoring area of ​​400m × 400m, using the coverage rate of sensor nodes as the optimization objective, obtain and sort the fitness values ​​of individual spiders; for example, the coverage matrix of individual spider k3. As shown:

[0061]

[0062] The fitness function of an individual spider k3, which is also known as coverage, is... As shown below, it is covered by at least one sensor.

[0063]

[0064] Step 3: Calculate the distance to individual spiders and update their positions. Assume the set of position coordinates for the target spider k3 is... Assume the set of location coordinates of spider k4 is Then the distance between the two spiders As shown below:

[0065]

[0066] The position update formula is:

[0067]

[0068] in Let i be the updated coordinates of the i-th sensor node of the target spider k3. Let x be the old coordinates of the i-th sensor node of the target spider k3 before the update. bestwhere a and β are the weight coefficients for controlling the exploration and exploitation of the spider, and rand() is a uniform random number generation function in the interval [0, 1].

[0069] where the social influence caused by other spiders is:

[0070]

[0071] Step 4: Propagation of the vibration signal of the spider individual, where the vibration signal intensity formula is as follows:

[0072]

[0073] where the fitness of the signal-producing spider k4 is: The distance between spider k4 and spider k3.

[0074] Step 5: Elite and exclusion strategies are adopted, and the positions of the remaining spiders are preserved. Finally, new spiders are randomly generated to join the population. In each iteration, elite spiders with high fitness are selected as elites according to the elite proportion Q (0 < Q < 1) and are not involved in exclusion. For individuals other than elite individuals, for example, the positions of spider 3 and spider 4 are

[0075]

[0076] The similarity is where C is the number of corresponding elements that are the same in the positions of the two spiders. Let the exclusion threshold be SimMAX (0 ≤ SimMAX ≤ 1). The spider individual with the highest exclusion degree is excluded, and a new spider is used to replace it. For example, the spider that satisfies Sim(k3, k4) > SimMAX will be excluded.

[0077] Step 6: The elite individuals and the remaining individuals after the exclusion operation are combined to form a new generation population. The new generation population will inherit the excellent genes from the previous generation population and maintain a certain diversity.

[0078] Step 7: Determine whether the set elite exclusion spider algorithm iteration threshold is reached, i.e., the maximum number of iterations is reached, the coverage rate reaches the preset threshold, or the required solution is found. When the algorithm threshold is reached, the algorithm ends, and the optimal deployment scheme is output. If the algorithm threshold is not reached, the fitness of the newly generated spider is calculated, and the attraction of the spider is recalculated and sorted in order. The above operations are repeated.

[0079] Through simulation and analysis, the mutation rate of the genetic algorithm in the comparison algorithm is 0.1, and the population size and the number of nodes are the same as those of the improved social spider algorithm. Figure 2As shown, by comparing the number of iterations and the proportion of coverage, the WSN target coverage optimization method of the improved swarm intelligence algorithm (improved social spider algorithm) compared with the genetic algorithm, the sensor node coverage rate of the method is effectively improved, and it tends to converge at the 43th generation, the target coverage rate after convergence is 95%, the running convergence speed is faster and the optimization ability is stable. The genetic algorithm compared in 83 generations tends to converge and the coverage value is 79%, the target coverage rate is much lower than the improved social spider algorithm.

[0080] Although the present application has been described in terms of specific embodiments, it will be apparent to those of ordinary skill in the art upon attaining an understanding of the foregoing description that numerous modifications and variations can be made therein without departing from the scope and spirit of the application. Thus, it is intended that the present application covers the modifications and variations of this application provided they come within the scope of the appended claims and their equivalents.

Claims

1. A WSN target coverage optimization method based on an improved swarm intelligence algorithm, characterized in that, It includes the following steps: First, construct and initialize the WSN target coverage model. Subsequently, taking the coverage rate of the WSN target area as the solution goal, construct a fitness function, and solve the fitness value of each individual in the spider population in the improved social spider algorithm through this function; Then, perform distance measurement and position update operations on each individual in the spider population, and at the same time realize the transmission of vibration signals between spider individuals. On this basis, adopt the elite and exclusion strategies to screen out the elite spider individuals with higher fitness and retain their positions, eliminate the objects to be excluded among the non-elite individuals, and randomly generate new spider individuals to supplement the population; Finally, judge whether the set iteration threshold of the improved social spider algorithm is reached. If not, loop the above operations. If the threshold is reached, output the position of the optimal spider individual and the optimal fitness value in the population, and determine the sensor position matrix corresponding to this optimal spider individual as the final optimal deployment plan of the WSN coverage optimization model.

2. The WSN target coverage optimization method based on the improved swarm intelligence algorithm according to claim 1, characterized in that, Construct the target coverage model and initialize it, including: Set the sensor node set \(F = \{f_1, f_2, \ldots, f i , \ldots, f n \}\), where \(f i \) represents the \(i\)-th sensor, and \(0 < i < n\). The monitored target set \(E=\{e_1, e_2, \ldots, e j , \ldots, e m \}\), where \(e j \) represents the \(j\)-th monitored target, and \(0 < j < m\); Define the sensing range \(R s \) and communication range \(R c \) of the sensor nodes, and \(R s < R c . Adopt a two-dimensional coordinate system to determine the two-dimensional coordinates of the sensor node \(f i \) as The position coordinates of the monitored target \(e j \) are And construct the coverage relationship matrix \(R\) between the sensor nodes and the monitored targets as shown below: Among the elements This means that the j-th monitored target was not detected by the f-th target. i Coverage of sensor nodes, Then it means that the fth i The sensor nodes cover the area, meaning the distance between them is less than R. s .

3. The WSN target coverage optimization method based on the improved swarm intelligence algorithm according to claim 1, characterized in that, The calculation of the fitness function includes the following steps: Select the size of the target area, the number of sensors, and the sensor sensing radius, determine the corresponding relationship between the target coverage concept and the improved social spider algorithm concept, and randomly initialize the positions of spider individuals in the monitoring area; among them, each set of sensor deployment plans corresponds to the corresponding spider individual, and the plan coverage rate corresponds to the spider individual fitness value; Using sensor node coverage as the optimization objective, the fitness values ​​of individual spiders are obtained and ranked; where the individual spider K... O Covering matrix As shown below: Set the set of sensor nodes \(F = \{f_1, f_2, \ldots, f i , \ldots, f n \}\), where \(f i \) represents the \(i\)-th sensor, and \(0 < i < n\). Among them, the sensing radius of the \(f i \)-th sensor node is \(r i \). The set of monitored targets \(E = \{e_1, e_2, \ldots, e j , \ldots, e m \}\), where \(e j \) represents the \(j\)-th monitored target, and \(0 < j < m\). The total number of sensor nodes is \(n\). represents that in the \(k O \)-th scheme, the \(i\)-th sensor node covers the \(j\)-th monitored target. represents not covered. Randomly initialize N spiders, representing N groups of sensors, and each group of sensors contains n sensor nodes; Spider k O Position: where 0 < O ≤ N; the two-dimensional coordinates of the ith sensor node among the k O spider individuals are The two-dimensional coordinates of the jth monitored target are Spider K o fitness function As shown below: in This represents the total number of targets covered by at least one sensor.

4. The WSN target coverage optimization method based on the improved swarm intelligence algorithm according to claim 1, characterized in that, The calculation of the distance of spider individuals and the update of spider positions include the following steps: Set the sensor node set F = {f1, f2, …, f i , … f n}, where f i represents the i-th sensor, and 0 < i < n. The position coordinate set of the target spider k t is The position coordinate set of other single spiders k v is where the total number of sensor nodes is n, 0 < t ≤ N, 0 < v ≤ N, N is the number of randomly initialized spiders, representing N groups of sensors, and each group of sensors contains n sensor nodes; k t The two-dimensional coordinates of the i-th sensor node in a group of spider individuals are: For Spider K t The x-axis coordinate of the i-th sensor node For Spider K t The y-axis coordinate of the i-th sensor node; To measure the distance between two spiders, consider the positions of all sensors. As shown below: The position update formula is: in For target spider k t The updated coordinates of the i-th sensor node. For target spider k t The old coordinates of the i-th sensor node before the update, x best Let α be the position of the i-th sensor corresponding to the spider with the highest fitness, α and β be the weight coefficients controlling the spider's exploration and development, and rand() be the uniform random number generation function in the interval [0,1]. Among them, the influence socialinfluence caused by other spiders on it is: Where V represents other spiders k v For target spider k t The intensity of the vibration signal.

5. The WSN target coverage optimization method based on the improved swarm intelligence algorithm according to claim 1, characterized in that, The vibration signal intensity V, and its formula is as follows: Spider k generates a signal v Adaptability, For Spider K v k t The distance between, k t For the target spider, k v For other individual spiders.

6. The WSN target coverage optimization method based on the improved swarm intelligence algorithm according to claim 1, characterized in that, Adopt the elite operator, exclusion strategy and retain the positions of the remaining spiders. Finally, randomly generate new spiders to join the group, including the following steps: Select the spider individuals with high fitness as elites according to the elite ratio Q (0 < Q < 1) and they will not be excluded; the total number of sensor nodes is n; Let the positions of spider k1 and spider k2 be: Define the sensor node set F = {f1, f2, ..., f i ,…f n }, where f i Let i represent the i-th sensor, and 0 <i<n, Let x be the x-coordinate of the i-th sensor node of spider k1. Let y be the y-coordinate of the i-th sensor node of spider k1; Let x be the x-coordinate of the i-th sensor node in spider k2. Let be the y-coordinate of the i-th sensor node of spider k2; Then there is similarity: Among them, C is the number of corresponding elements that are the same in the positions of the two spiders; Calculate the similarity Sim(k1,k2) of two individuals, where k1 and k2 are the position vectors of the two spiders respectively; Set the exclusion threshold: SimMAX (0 ≤ SimMAX ≤ 1), and exclude the spider individual with the highest exclusion degree, and replace it with a new spider: In the formula, PAI represents the exclusion operation, and the spider with low fitness among the spiders that satisfy PAI = 1 will be excluded and replaced with a randomly generated new spider.

Images