A WSN Topology Control Method Based on Dynamic Density and Cooperative Sensing

Abstract

The present invention discloses a WSN topology control method based on dynamic density and cooperative perception, which is used to isolate the harm that malicious nodes within a cluster may cause. The method includes: uniformly deploying trust evaluation nodes in the network, mutually evaluating the trust evaluation nodes, sending the evaluation results to their respective corresponding cluster heads, routing through the cluster heads to the sink node, and the sink node calculates the trust values of the evaluation nodes and decides whether to update the evaluation nodes. The present invention makes certain improvements to the trust evaluation model, combines the current trust value and historical trust value of the node, as well as the remaining energy of the node, dynamically reflects the change of the node trust value, enables the protocol to maintain a relatively high and stable packet transmission rate when suffering from selective forwarding attacks, and at the same time extends the network lifetime.

Description

Technical Field

[0001] The present invention belongs to the field of routing security of wireless sensor networks, and particularly relates to a topology control method for wireless sensor networks based on dynamic density and cooperative perception. Background Art

[0002] Wireless Sensor Networks (WSN) is a distributed sensing network, which is a wireless network composed of a large number of stationary or mobile sensors in an ad-hoc and multi-hop manner. Currently, it is widely used in military, environmental monitoring, medical, industrial production, traffic control and other fields. However, due to limitations such as the computing power, storage capacity, and energy of nodes, the development of wireless sensor networks has been affected. In hierarchical routing protocols, trust evaluation nodes are usually served by cluster head nodes. In order to better balance the network energy consumption, some scholars consider decentralizing the trust evaluation work of the cluster head to other nodes within the cluster or the sink node.

[0003] Early on, Wu et al. proposed the idea of selecting a trustworthy node within the cluster as a substitute for the cluster head. This substitute cluster head performs the trust evaluation work instead of the cluster head, and then the cluster head can refuse to receive data from nodes with insufficient trust during intra-cluster communication, isolating the potential harm caused by malicious nodes within the cluster.

[0004] Wei et al. proposed a mechanism in which a dedicated node serves as a trust cloud to supervise network security based on the work of Cai et al. In this mechanism, the sink node divides the sensing network based on the K-means clustering algorithm, and then selects the cluster head and the trust cloud node within each cluster according to the cluster head selection function and the trust evaluation node selection function. The trust cloud node defines the trust value of the node according to the similarity between the trust vector of each node and its own trust vector. However, this mechanism has the problem of unbalanced energy consumption of trust evaluation nodes, which may cause the trust evaluation nodes to die earlier when there are more nodes within the cluster.

[0005] In summary, the design of security topology control methods has become a research hotspot in wireless sensor networks in recent decades. How to balance the energy consumption of nodes and ensure the transmission security of data has become an urgent problem in this field. Summary of the Invention

[0006] In order to solve the technical problems mentioned in the above background art, the present invention proposes a WSN topology control method based on dynamic density and cooperative perception.

[0007] In order to achieve the above technical objectives, the technical solution of the present invention is as follows:

[0008] A WSN topology control method based on dynamic density and cooperative perception, comprising the following steps:

[0009] (1) It is stipulated that nodes compete for cluster heads according to their own competition values and competition radii. After a cluster is determined, the nodes in this cluster will terminate the process, as shown in Figure 1 the following;

[0010] (2) In the trust evaluation mechanism, the sink node first divides the network into several regions of the same size, and then analyzes the location, energy, and trust value information of nodes in each region based on the grey clustering analysis method, and selects the nodes belonging to the high-quality grey class among them to act as trust evaluation nodes;

[0011] (3) The sink node updates the trust values of the evaluation nodes according to the mutual evaluation results among the trust evaluation nodes, combined with the methods in statistics, and decides whether to reselect the evaluation nodes according to the trust values.

[0012] Furthermore, in step (1), the method for constructing network clusters is as follows:

[0013] (101) The sink node calculates the node density d of each node according to the distribution of the neighboring nodes of the i-th node x in the sensor network i and the distribution of nodes around the neighboring nodes. The calculation formula is as follows: i

[0014]

[0015] where m is the number of nodes in the neighboring node set nbor(i) of node x i , n is the number of nodes in the neighboring node set nbor(j) of the j-th node in nbor(i), and d ij represents the Euclidean distance from node x i to the j-th node in nbor(i);

[0016] (102) Combining the remaining energy E i and trust value T i of the node, calculate its competition value y i . The calculation process of the competition value is as follows:

[0017]

[0018] (103) After the competition values of all nodes are determined, calculate the competition radius r i of each node:

[0019]

[0020] where N represents the total number of surviving nodes in the network;

[0021] (104) Select the node with the largest competition value as the cluster head node. Nodes within the competition radius of this cluster head node join the cluster and become its member nodes;

[0022] (105) Nodes that have already formed clusters exit the network clustering process, and the remaining nodes update the node density and repeat (101) to (104);

[0023] (106) Repeat the above steps until the number of cluster head nodes in the network reaches 4% of the total number of nodes, and the network clustering stage ends.

[0024] Further, in step (2), the deployment method of the trust evaluation nodes is as follows:

[0025] (201) Obtain the abscissas and ordinates of all nodes in the area to form an abscissa list and an ordinate list;

[0026] (202) Calculate the center point coordinates, that is, use the mean of the abscissa list and the mean of the ordinate list as the center point coordinates of this area. Let the number of nodes in the area be M, and the calculation method is as follows:

[0027]

[0028] where x i and y i represent the abscissa and ordinate of the i-th node x i respectively.

[0029] (203) Calculate the position attribute of each node according to the distance of the node from the center point, and the calculation method is:

[0030]

[0031] (204) Design the observation attributes as the remaining energy, trust value, and position attribute, design "excellent", "medium", and "poor" as three gray classes to represent the degree to which a node is good enough to act as a trust evaluation node, and design the possibility function;

[0032] (205) Calculate the weights of each attribute according to the possibility function;

[0033] (206) According to the remaining energy, trust value, and position attribute of each node, use the gray clustering evaluation algorithm to calculate the clustering coefficient of the node belonging to the gray class "excellent";

[0034] (207) The node with the highest coefficient in each area becomes the evaluation node.

[0035] Further, in step (3), the method for evaluating the trust degree of nodes within the cluster is as follows:

[0036] (301) Within each network sub-region, the trust evaluation node evaluates the trust values of all nodes in the region. If there are trust evaluation nodes in the upper, lower, left, and right regions adjacent to this trust evaluation node, then this trust evaluation node also evaluates the trust of the evaluation nodes in the upper, lower, left, and right regions.

[0037] (302) The trust value of a node is jointly determined by its current and historical behaviors.

[0038] (303) After the aggregation node receives the data from each trust evaluation node, it creates a two-dimensional list. For each trust evaluation node, according to the trusted value provided by its neighbor trust evaluation nodes, it calculates the trust value of the trust evaluation node using the absolute median difference method in statistics.

[0039] Beneficial effects brought by adopting the above technical solutions:

[0040] (1) The present invention is coordinated by the base station for operation, and trust evaluation nodes are evenly deployed in the network, so that the energy consumption of the trust evaluation nodes in the network remains relatively consistent. At the same time, the trust evaluation nodes evaluate each other, send the evaluation results to their respective corresponding cluster heads, and route them to the aggregation node through the cluster heads. The aggregation node calculates the trust value of the evaluation node and decides whether to update the evaluation node. Through the cooperation of the evaluation node and the aggregation node, the secure perception of the whole network state is realized, the accuracy of discovering malicious nodes is improved, and the misjudgment rate is reduced.

[0041] (2) Different from other trust evaluation mechanisms, the selection of the evaluation node in the present invention adopts the grey clustering evaluation algorithm, and the aggregation node uses the absolute median difference to calculate and update the trust value of the evaluation node. Combining the current trust value and historical trust value of the node, as well as the remaining energy of the node, it dynamically reflects the change of the node's trust value, making the evaluation result more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is the network clustering flowchart of the present invention;

[0043] Figure 2 is the schematic diagram of node distribution of the present invention; DETAILED DESCRIPTION OF THE INVENTION

[0044] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0045] A topology control method for a low-power wireless sensor network based on trust evaluation includes the following steps:

[0046] Step 1: It is stipulated that a node runs for the cluster head according to its own competition value and competition radius. When a clustering is determined, the nodes within this cluster terminate the process.

[0047] Step 2: In the trust evaluation mechanism, the sink node first divides the network into several regions of the same size, and then analyzes information such as the location, energy, and trust value of nodes in each region based on the grey clustering analysis method, and selects the nodes belonging to the high-quality grey class as trust evaluation nodes;

[0048] Step 3: The sink node updates the trust value of the evaluation nodes according to the mutual evaluation results among the trust evaluation nodes, combined with the method in statistics, and decides whether to reselect the evaluation nodes according to the trust value;

[0049] In this embodiment, the above Step 1 can be implemented by the following preferred scheme:

[0050] 101. The sink node calculates the node density d of each node according to the distribution of neighboring nodes of the i-th node x in the sensor network and the distribution of nodes around the neighboring nodes. The calculation formula is as follows: i where m is the number of nodes in the neighboring node set nbor(i) of node x, n is the number of nodes in the neighboring node set nbor(j) of the j-th node in nbor(i), and d represents the Euclidean distance from node x to the j-th node in nbor(i); i The calculation formula is as follows:

[0051]

[0052] where m is the number of nodes in the neighboring node set nbor(i) of node x i n is the number of nodes in the neighboring node set nbor(j) of the j-th node in nbor(i), and d represents the Euclidean distance from node x to the j-th node in nbor(i); ij represents the Euclidean distance from node x i to the j-th node in nbor(i);

[0053] 102. Combining the remaining energy E of the node, the initial energy E0 and the trust value T of the node, calculate its competition value y. The calculation process of the competition value is as follows: i The calculation process of the competition value is as follows: i The calculation process of the competition value is as follows: i The calculation process of the competition value is as follows:

[0054]

[0055] 103. After the competition values of all nodes are determined, calculate the competition radius r of each node: i :

[0056]

[0057] where N represents the total number of surviving nodes in the network;

[0058] 104. Select the node with the largest competition value as the cluster head node, and the nodes within the competition radius of this cluster head node join the cluster and become its member nodes;

[0059] 105. The nodes that have already formed a cluster exit the network clustering process, and the remaining nodes update the node density and repeat 101 to 104;

[0060] 106. Repeat the above steps until the number of cluster head nodes in the network reaches 4% of the total number of nodes, and the network clustering stage ends.

[0061] In this embodiment, the above step 2 can be implemented by adopting the following preferred solution:

[0062] 201. Obtain the horizontal coordinates and vertical coordinates of all nodes in the region to form a horizontal coordinate list and a vertical coordinate list;

[0063] 202. Calculate the coordinates of the center point, that is, take the mean of the horizontal coordinate list and the mean of the vertical coordinate list as the coordinates of the center point of the area. Suppose the number of nodes in the area is M. The calculation method is as follows:

[0064]

[0065] where x i and i Represents the i-th node x i The horizontal and vertical coordinates of .

[0066] 203. According to the distance between the node and the center point, the position attribute of each node is calculated as follows:

[0067]

[0068] 204. Design the observed attributes as residual energy, trust value and location attributes, design "excellent", "medium" and "poor" as three gray categories to represent whether the node is good enough to serve as a trust evaluation node, and design the possibility function;

[0069] 205. Calculate the weight of each attribute according to the possibility function;

[0070] 206. According to the residual energy, trust value and location attribute of each node, the clustering coefficient of the node belonging to the gray class "excellent" is calculated using the gray clustering evaluation algorithm;

[0071] 207. The node with the highest coefficient in each region becomes the evaluation node.

[0072] In this embodiment, the above step 3 can be implemented by adopting the following preferred solution:

[0073] 301. In each network sub-area, the trust evaluation node evaluates the trust values of all nodes in the area. In addition, if there are trust evaluation nodes in the four areas above, below, left, and right of the trust evaluation node, the trust evaluation node also performs trust evaluation on these evaluation nodes.

[0074] 302. The trust value of a node is jointly determined by its current and historical behaviors. During the trust evaluation process, the evaluating node records the number of data packets sent, m1, and the number of data packets received, m2, between itself and the evaluated node, and uses the beta distribution model to calculate the trust value. The calculation method is as follows:

[0075]

[0076] 303. To defend against attack methods with discontinuous attack times such as switch attacks, the evaluation mechanism includes recording the historical behaviors of nodes. Therefore, the evaluating node uses the ratio of the number of historical successful interactions to the number of historical failed interactions of each node as a coefficient to calculate the trust value of the node. The calculation method is as follows:

[0077]

[0078] Where malnum represents the number of historical failed interactions, cnum represents the number of historical successful interactions, and α is a coefficient introduced to ensure the validity of the formula. In one interaction behavior, if the number or length of the data packets returned by the evaluated node is different from the number or length of the data packets sent by the evaluating node, then this interaction is considered a failed interaction.

[0079] 304. After the sink node receives the data from each trust evaluation node, it establishes a two-dimensional list. For each trust evaluation node, according to the trusted value provided by its neighbor trust evaluation nodes, it uses the absolute median difference method in statistics to calculate the trust value of the trust evaluation node.

Claims

1. A WSN topology control method based on dynamic density and collaborative awareness, characterized in that The following steps are involved: (1) Nodes are required to compete for cluster heads based on their own competition values and competition radius. Once a cluster is determined, the nodes in the cluster terminate the process. The specific steps are as follows: (101) The sink node calculates the node density d of each node according to the distribution of neighboring nodes of the ith node x in the sensor network and the distribution of nodes around the neighboring nodes. i The calculation formula is as follows: i ​ where m is the number of nodes in the neighbor node set nbor(i) of node x i , n is the number of nodes in the neighbor node set nbor(j) of the j-th node in nbor(i), and d ij represents the Euclidean distance from node x i to the j-th node in nbor(i); (102) Combine with the density d of the node itself i , the remaining energy E i , the initial energy E0 and the trust value T of the node i , calculate its competition value y i , the calculation process of the competition value is as follows: (103) After the competition values of all nodes are determined, calculate the competition radius r of each node i : Where N represents the total number of surviving nodes in the network; (104) Select the node with the largest competition value as the cluster head node, and the nodes within the competition radius of the cluster head node join the cluster and become its member nodes; (105) The clustered nodes exit the network clustering process, and the remaining nodes update their node density, and repeat (101) to (104); (106) Repeat the above steps until the number of cluster head nodes in the network reaches 4% of the total number of nodes, and the network clustering stage ends; (2) In the trust evaluation mechanism, the sink node divides the network into several areas of equal size. In each area, the location, energy, and trust value information of the nodes are analyzed based on the gray clustering analysis method, and the nodes belonging to the high-quality gray class are selected to serve as trust evaluation nodes. The specific steps for the deployment of trust evaluation nodes are as follows: (201) Obtaining the horizontal coordinates and vertical coordinates of all nodes in the region to form a horizontal coordinate list and a vertical coordinate list; (202) Calculate the coordinates of the center point, that is, take the mean of the horizontal coordinate list and the mean of the vertical coordinate list as the coordinates of the center point of the region, and assume that the number of nodes in the region is M. The calculation method is as follows: where x i and y i represent the abscissa and ordinate of the i-th node x i respectively; (203) According to the distance between the node and the center point, the position attribute of each node is calculated as follows: (204) Design the observed attributes as residual energy, trust value and location attributes, design "excellent", "medium" and "poor" as three gray categories to represent whether the node is good enough to serve as a trust evaluation node, and design the possibility function; (205) Calculate the weight of each attribute according to the possibility function; (206) According to the residual energy, trust value and location attribute of each node, the clustering coefficient of the node belonging to the gray class "excellent" is calculated using the gray clustering evaluation algorithm; (207) The node with the highest coefficient in each region becomes the evaluation node; (3) The aggregation node updates the trust value of the evaluation node based on the mutual evaluation results between the trust evaluation nodes and combines statistical methods, and decides whether to reselect the evaluation node based on the trust value. The specific steps are as follows: (301) In each network sub-area, the trust evaluation node evaluates the trust values of all nodes in the area. If there are trust evaluation nodes in the upper, lower, left, and right areas immediately adjacent to the trust evaluation node, the trust evaluation node also performs trust evaluation on the evaluation nodes in the upper, lower, left, and right areas. (302) The trust value of a node is determined by its current and historical behaviors. During the trust evaluation process, the evaluating node records the number of packets sent m1 and received m2 between the evaluating node and the evaluated node, and uses the β distribution model to calculate the trust value. The calculation method is as follows: (303) After receiving the data from each trust evaluation node, the aggregation node creates a two-dimensional list. For each trust evaluation node, the trust value of the trust evaluation node is calculated using the absolute median difference method in statistics based on the trusted values provided by its neighbor trust evaluation nodes.

Images