Reliability Evaluation Method for Wireless Sensor Network Coverage Based on Trusted Information Coverage

Abstract

The present invention discloses a method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage: deploying a wireless sensor network according to a monitoring scenario; comprehensively considering the polymorphism of nodes, the connectivity between nodes, the network coverage rate, and the root mean square error to define the trusted information coverage reliability; evaluating the state transition of sensor nodes based on a three-state node model; constructing a coverage table structure to uniformly describe the coverage reliability of each node in the network; evaluating the coverage reliability of each sub-region by setting virtual reconstruction nodes and using the coverage tables of the virtual reconstruction nodes; calculating the overall network coverage rate and the lower bound of the coverage reliability by means of table combination. The present invention defines the coverage of sensors from the perspective of information collaboration, fully utilizes the spatial correlation of detection variables, and increases the coverage area of the network. The coverage reliability of each node in the network is uniformly described using a coverage table, and the lower bound of the overall network coverage reliability is efficiently calculated by means of coverage table combination.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless sensor networks, and more specifically, relates to a method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage. Background Art

[0002] As a powerful information acquisition technology, wireless sensor networks (WSNs) have been widely deployed in various practical applications. The coverage reliability problem quantifies the ability of a WSN to provide a specific sensing coverage requirement and successfully transmit the sensed data to the sink node. Higher coverage reliability can ensure the information acquisition and transmission of a WSN, thereby improving the quality of service (QoS). However, due to the characteristics of wireless sensor networks themselves and the diversity and particularity of the application environment, the operation of the network will be affected by adverse factors such as the number of nodes, random node failures, communication link eavesdropping, and attacks, which may lead to local or even overall network failure, and the coverage ability of the network cannot meet the application requirements. Therefore, evaluating the coverage reliability of the network before network deployment is an important task.

[0003] The difficulties in evaluating the coverage reliability of wireless sensor networks are mainly reflected in three aspects. First, the coverage reliability is closely related to the coverage model that describes the sensor coverage ability. Most of the past research on coverage reliability was based on the disk coverage model, but the disk model is too idealized and simplistic and does not fit the deployment requirements of actual applications. Second, for a WSN composed of k three-state sensors, comprehensively considering all 3 k possible network topology states to accurately calculate the reliability is a #P-hard problem, and there is unlikely to be an exact reliability solution for many network topologies. Third, most of the existing methods only consider coverage reliability or connection reliability separately, reducing the comprehensiveness of the reliability evaluation. Summary of the Invention

[0004] Aiming at the above defects or improvement requirements of the prior art, the present invention provides a method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage, aiming to comprehensively and effectively evaluate the coverage reliability of a wireless sensor network and assist engineers and technicians in designing a more reliable wireless sensor network. The reliability evaluation method defines the sensor coverage range from the perspectives of prediction and information reconstruction using the trusted information coverage model, fully utilizes the spatial correlation of the monitored physical parameters and the cooperation between adjacent nodes, uses virtual reconstruction nodes to uniformly describe the coverage reliability of each sub-region, uses the connection strength matrix to describe the connection reliability between nodes, and efficiently evaluates the coverage reliability of the network by constructing a virtual network and combining coverage table operations, improving the evaluation efficiency and application scope, thereby solving the problem of reliability evaluation of wireless sensor networks and having good guiding and reference value.

[0005] To achieve the above object, according to one aspect of the present invention, a method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage is provided, including the following steps:

[0006] (1) Establish a network model according to the monitoring coverage scenario of the target area;

[0007] (1.1) Determine the number, location, sensing radius, communication radius, working probability P sens of the sensor sensing module, and the working probability P com of the sensor communication module. Based on the operation process of the three-state sensor node model, there are the following states: cs represents the intact operation state, c represents the relay state of only communication, and f represents the completely failed state. The working probabilities of different states are:

[0008] P cs (v)=P com (v)×P sense (v)

[0009] P c (v)=P com (v)×(1 - P sense (v))

[0010] P f (v)=1 - P cs (v)-P c (v)=1 - P com (v)

[0011] (1.2) Model the wireless sensor network WSN as an undirected graph W = {{v sinj} ∪ V, E};

[0012] (1.3) According to the spatial correlation of the monitoring variables, set appropriate range CR and root mean square error RMSE. According to the range, divide the target area Δ = Δ l × Δ w into several grid regions G = {g1, g2, g3,..., g n}, and the center of each grid region is the reconstruction point;

[0013] (2) Construct the coverage table of each sensor node according to the polymorphism of the sensor nodes, denoted as C V ;

[0014] (3) Deploy virtual reconstruction nodes VRN in each grid region to construct a virtual network; define the coverage reliability of the wireless sensor network based on trusted information coverage;

[0015] (4) Calculate the coverage table of VRN based on the trusted information coverage model;

[0016] (5) Calculate a reliable coverage table aggregation path according to the connectivity status among VRNs;

[0017] (6) Perform a combination operation on the coverage tables of VRNs one by one according to the coverage table aggregation path until all the coverage table information is combined into the coverage table of the VRN sink node;

[0018] In one embodiment of the present invention, the method for constructing the coverage table in step (2) is as follows:

[0019] For any multi-state sensor node v ∈ V in the network, its state set is represented as a sense (v u ) represents the sensing area of the sensor node v in the u state, then the possible area values that the sensor node v can sense are:

[0020]

[0021] a sense (v U ) = {a1, a2, a3,..., a n}

[0022] The structure of the coverage table C v of the sensor node v is represented as:

[0023] C v [a i = ∑Prob(U v : a sense (v U ) = a i )(a i > 0)

[0024] C v [0] = P c (v) = 1 - ∑C v [a i - P f (v)

[0025] In one embodiment of the present invention, the virtual reconstruction node in step (3) refers to a virtual sensor node set at the reconstruction point position in each grid area. The virtual reconstruction node is similar to an ordinary sensor, has multiple working states and has a coverage table C VRN describing the node coverage ability, and its working state is jointly determined by all the nodes in the grid, and its coverage table describes the coverage reliability of the grid area where it is located.

[0026] In one embodiment of the present invention, the virtual network in step (3) refers to a virtual wireless sensor network constructed by VRN nodes, denoted as WVRN = ({VRN sink} ∪ VRB, VM), where VRN sink represents the virtual sensor nodes in the grid where the sink node is located, and VM represents the set of communication links between VRNs.

[0027] In one embodiment of the present invention, the definition of coverage reliability (CACREL) based on the trusted information coverage model in step (3) means that for a WSN deployed in area S square , A req is the coverage rate threshold required by the sensing service, representing the percentage of the minimum coverage area. Then CACREL refers to the probability that there exists a subset of cs-state nodes such that the data stream oriented to CIC generated by V′ itself that meets the sensing service requirements can reach the sink node.

[0028] In one embodiment of the present invention, in step (3), according to step (2), it can be calculated through the coverage table of the sink node whether the entire wireless sensor network can meet the sensing service requirements. After converting the ordinary WSN into a virtual WSN, the calculation formula for the CACREL of the network is:

[0029] S min = S square × A req

[0030]

[0031] In one embodiment of the present invention, step (4) specifically includes the following sub-steps:

[0032] (4.1) Calculate the trusted information coverage in different working states within each grid area;

[0033] (4.2) Calculate the probability that the working state that meets the trusted information coverage requirements occurs; for any working state the probability that the network operates in this state is calculated by the following formula:

[0034]

[0035] where V g is the set of sensor nodes located in grid g, is the set of cs-state sensors, is the set of c-state sensors, is the set of f-state sensors.

[0036] (4.3) Construct the coverage table C of VRNVRN ; The a of each VRN sense (VRN) includes two numerical values as follows:

[0037]

[0038] Wherein, S g represents the area of the grid, and the area value of the grid is usually CR×CR m 2 , but if the field cannot be precisely divided by CR, the area of some grids will be less than CR×CR m 2 .

[0039] The coverage table C of VRN VRN contains 2 entries as follows:

[0040]

[0041] C VRN [0] = 1 - C VRN [S g - P f (VRN)

[0042] In an embodiment of the present invention, the step (4.1) specifically includes the following sub-steps:

[0043] (4.1.1) If the grid g contains y three-state sensor nodes, the set of possible working states in this grid area is:

[0044]

[0045] (4.1.2) Enumerate all possible working states, and use the sensor nodes in the cs state to perform collaborative information reconstruction on the perception data of the reconstruction points in each working state, and calculate the root mean square error RMSE.

[0046] (4.1.3) In the credible information coverage model, for the reconstruction point x i , use the ordinary Kriging interpolation function to calculate the estimated value of the environmental variable of the reconstruction point x i , that is, use the weighted average of the measurement values of the sensor nodes in the cs state within the reconstruction neighborhood Z(x i ) to calculate the estimated value of the environmental variable; the interpolation weight coefficient ω i of the sensor nodes within the neighborhood satisfies is the number of sensor nodes v i within the reconstruction neighborhood Z(x i );

[0047] (4.1.4) Combine the ordinary Kriging interpolation function to calculate the root mean square error Φ(x) of the reconstruction point x, and the calculation expression is: Among them and μ(x) are solved by the following formula;

[0048] (4.1.5) Interpolation weight coefficient λ i A set of optimal solutions is obtained through the minimum Kriging variance; the Lagrange multiplier μ(x) is introduced to generate a linear Kriging system composed of n + 1 equations with n + 1 unknowns, and the interpolation weight coefficient λ is obtained after solving i ;

[0049]

[0050] Among them, γ(v i , v j ) and γ(v i , x) are calculated through the variogram;

[0051] (4.1.6) Calculate γ(v i , v j ) and γ(v i , x) in step (4.1.5); select the Gaussian variogram as the variogram of the environmental variable to describe the spatial correlation between the data collected by the sensor node v i ; the formula of the Gaussian variogram is:

[0052]

[0053]

[0054] Among them, is the Euclidean distance between the sensor node v i and the reconstruction point x, is the Euclidean distance between the sensor node v i and v j , and both C0 and C1 are constants;

[0055] (4.1.7) According to the definition of the credible information coverage model, if that is, the time-averaged root mean square error is greater than the set coverage threshold, then this grid area meets the credible information coverage requirement in this working state, otherwise it is not covered, and this invalid working state is ignored and the enumeration continues;

[0056] In one embodiment of the present invention, step (5) specifically includes the following sub-steps:

[0057] (5.1) Calculate VM; for the network W = { { v sink} ∪ V, E}, where the communication link e ij ∈E is defined as:

[0058]

[0059] Given two adjacent grids g a and g b , VM ab = k means that there are k links connecting the VRNs in g a and g b . VM ab is calculated by the following formula:

[0060]

[0061] (5.2) VM reflects the reliability of the communication connection between VRNs. The larger the value of VM ab , the stronger the connection between VRN a and VRN b . To make each C VRN converge to the sink node more reliably, we set the connection weight matrix h = 1 / VM to represent the selection priority of the links in VM. The smaller h is, the more likely this path is to be selected for coverage table aggregation.

[0062] (5.3) Based on the connection weight matrix h, use the dijkstra algorithm to generate a minimum spanning tree, and this tree-like path is the reliable coverage table aggregation path RPath;

[0063] In an embodiment of the present invention, the step (6) specifically includes the following sub-steps:

[0064] (6.1) Combined calculation of the coverage table; for two wireless sensor networks W1, W2, the two networks only share the same sink node, and the rest of the nodes are disjoint. Then the coverage table of the network composed of W1 and W2 is:

[0065]

[0066] where C1 is the coverage table of W1, C2 is the coverage table of W2, and C1×C2 represents the combined operation of the coverage tables; if the coverage table C i has n i coverage records, then C[m] has at most n1×n2 coverage records;

[0067] (6.2) Calculate the reliability evaluation value CACREL of the output network; according to the reliable coverage table aggregation path RPath calculated in (5.3), starting from the leaf nodes, perform the coverage table combination operation with its parent nodes one by one according to (6.1); after all leaf nodes have performed the combination operation, remove all leaf nodes; continuously perform the combination-removal operation until the coverage tables of all nodes are combined, and only the VRN sinkNode; the final reliability evaluation value CACREL of the entire network is obtained according to the CACREL calculation formula;

[0068] Generally speaking, compared with the prior art, the above technical solution conceived by the present invention has the following beneficial effects:

[0069] (1) High coverage reliability. The invention comprehensively explores the spatial correlation of the monitoring and reconstruction points in the covered target area from the perspective of information collaboration, and uses the root mean square error to estimate the coverage error to complete the coverage prediction, improve the coverage rate, and then improve the coverage reliability;

[0070] (2) Fast evaluation speed. In the invention, the coverage capabilities of each node are uniformly described by a coverage table. By calculating the reliability in partitions, the problem of complex calculation caused by full-state enumeration of network nodes is avoided, and the evaluation time is saved;

[0071] (3) Comprehensive evaluation indicators. The invention comprehensively considers the influence of factors such as multi-state nodes, link reliability, network coverage area, and coverage quality on the network coverage reliability, and can more comprehensively reflect the network coverage reliability;

[0072] (4) Strong versatility. The invention calculates the network coverage reliability by combining a virtual network and a coverage table, and is applicable to high-density large-range networks. In addition, the Internet of Things involved in the invention is a general-purpose network and is applicable to various application scenarios. The trusted information coverage model adopted in the invention can be used for coverage of different terrains, regions, and different data monitoring targets. Description of the Drawings

[0073] Figure 1 It is a flowchart of a method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage provided by an embodiment of the present invention;

[0074] Figure 2 It is a schematic diagram comparing the topological structures of a general network (left) and a virtual network (right) provided by an embodiment of the present invention; where Figure 2 (a) is a general network, Figure 2 (b) is a virtual network;

[0075] Figure 3 It is a schematic diagram of trusted information reconstruction within a grid area provided by an embodiment of the present invention;

[0076] Figure 4 It is a schematic diagram of the weight coefficients of each side after converting a general network (left) into a virtual network provided by an embodiment of the present invention (right); where Figure 4 (a) is a general network, Figure 4 (b) is a virtual network;

[0077] Figure 5Combined operations of the coverage table provided by the embodiments of the present invention. Detailed implementation manners

[0078] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0079] The following first explains and describes the technical terms of the present invention:

[0080] Correlation Range (CR): The distance critical value characterizing the spatial correlation of environmental variables. For a specific environmental variable and a spatial point, only the values of other spatial points within the correlation range are related to the current spatial point.

[0081] Root Mean Square Error (RMSE): Used to measure and evaluate the reconstruction and estimation quality of the spatial environmental variable values not adopted, that is, the error measure between the estimated value and the reference point value.

[0082] Confident Information Coverage: In the target monitoring area, if the root mean square error of the reconstruction information at a spatial point in this area is less than or equal to the threshold μ proposed by the actual application requirements, then this spatial point is covered by confident information.

[0083] Euclidean distance: Measures the absolute distance between two points or vectors in a multi-dimensional space, that is, the square root of the difference between vectors. The Euclidean distance from point A(x i , y i ) to point B(x j , y j ) is

[0084] Kriging interpolation: Kriging method is essentially a sliding weighted average method, with characteristics such as optimal, linear, and unbiased. The Kriging method is a regression algorithm for spatially modeling and predicting (interpolating) a random process or a random field based on the covariance function. In a specific random process, such as an intrinsically stationary process, the Kriging method can give an optimal linear unbiased estimate, so it is also called a spatial optimal unbiased estimator in geostatistics.

[0085] Adjacent node: Other sensor nodes whose Euclidean distance from the sensor node v is within its communication range R c are adjacent nodes of v.

[0086] Virtual reconstruction node: A virtual node set at the position of each grid reconstruction point, whose working state is jointly determined by all nodes within the grid, and its coverage table C VRN Describes the coverage reliability of the grid area where it is located.

[0087] Virtual network: Refers to a virtual wireless sensor network constructed by VRN nodes, denoted as W VRN =({VRN sink}∪VRN, VM).

[0088] Coverage table: A table constructed based on the polymorphism of sensor nodes to describe the coverage ability and working reliability of sensors.

[0089] Coverage table combination: For a new network formed by combining two wireless sensor networks that only share a sink node, its coverage table can be calculated through the coverage table combination operation of the two sub-networks.

[0090] Solutions to the difficulties existing in the prior art are as follows:

[0091] Regarding Difficulty 1, in the existing research on the coverage reliability of wireless sensor networks, the disk coverage model is mostly used to describe the coverage ability of sensor nodes, which is too simplistic and idealistic. Adopting a novel credible information coverage model can define the sensor coverage range from the perspectives of prediction and information reconstruction. The credible information coverage model makes full use of the spatial correlation of the monitored physical parameters and the collaborative cooperation of adjacent nodes, can meet the actual application requirements, and can reduce the number of required sensors under the same conditions. Regarding Difficulty 2, enumerating all network topology states of a wireless sensor network to calculate reliability is a #P-hard problem, and there is no exact reliability evaluation value for some topologies. Convert the network into a virtual network composed of virtual reconstruction nodes, use the coverage table of virtual nodes to accurately describe the coverage reliability of a sub-region, and then calculate the lower bound of the entire network coverage reliability through the combination operation of the coverage tables. Regarding Difficulty 3, in the existing research on the coverage reliability of wireless sensor networks, most only consider coverage reliability or connection reliability separately. Consider coverage and connectivity comprehensively, use the connection strength matrix to describe the communication reliability between virtual reconstruction nodes, and improve the reliability and accuracy of the coverage table combination operation.

[0092] As Figure 1 shown, the wireless sensor network coverage reliability evaluation method based on credible information coverage of the present invention includes the following steps:

[0093] (1) As Figure 2 (a) shown, establish a network model according to the monitoring coverage scenario of the target area;

[0094] (1.1) Determine the number, location, sensing radius, communication radius, and the working probability \(P\) of the sensing module of the sensor nodes to be deployed sense , and the working probability \(P\) of the communication module of the sensor com . Based on the operation process of the three - state sensor node model, there are the following states: \(cs\) represents the fully operational state, \(c\) represents the relay state with only communication, and \(f\) represents the completely failed state. The working probabilities of different states are:

[0095] \(P\) cs (v)= \(P\) com (v)× \(P\) sense (v)

[0096] \(P\) c (v)= \(P\) com (v)×(1 - \(P\) sense (v))

[0097] \(P\) f (v)=1 - \(P\) cs (v)- \(P\) c (v)=1 - \(P\) com (v)

[0098] (1.2) Model the wireless sensor network WSN as an undirected graph \(W=\{\{V\) sink}\cup V, E\}\);

[0099] (1.3) According to the spatial correlation of the monitored variables, set appropriate range \(CR\) and root - mean - square error \(RMSE\). Based on the range, divide the target area \(\Delta=\Delta\) l ×\(\Delta\) w into several grid regions \(G = \{g1, g2, g3,\cdots, g\) n}\), and the center of each grid region is the reconstruction point;

[0100] (2) For any multi - state sensor node \(v\in V\) in the network, its state set is represented as a sense (v u ) represents the sensing area of sensor node \(v\) in the \(u\) state. Then the possible sensing area values of sensor node \(v\) are:

[0101]

[0102] a sense (v U )=\{a1, a2, a3\cdots, a\) n}\}

[0103] The coverage table \(C\) v of sensor node \(v\) is structured as:

[0104] \(C\) v[a i = ∑ Prob(U v : a sense (v U ) = a i ) (a i > 0)

[0105] C v [0] = P c (v) = 1 - ∑ C v [a i - P f (v)

[0106] (3) As shown in Figure 2 (b), deploy virtual reconstruction nodes (VRN) in each grid area to construct a virtual network; define the coverage reliability of a wireless sensor network based on trusted information coverage;

[0107] A virtual reconstruction node refers to a virtual sensor node set at the reconstruction point position in each grid area. A virtual reconstruction node is similar to an ordinary sensor, has multiple working states, and has a coverage table C VRN that describes the node's coverage ability. Its working state is jointly determined by all the nodes in the grid, and its coverage table describes the coverage reliability of the grid area where it is located.

[0108] A virtual network refers to a virtual wireless sensor network constructed by VRN nodes, denoted as W VRN = ({VRN sink}} ∪ VRN, VM), where VRN sink represents the virtual sensor node in the grid where the sink node is located, and VM represents the set of communication links between VRNs.

[0109] The definition of the coverage reliability CACREL based on the trusted information coverage model means that for a WSN deployed in area S square , A req is the coverage rate threshold required by the sensing service, representing the percentage of the minimum coverage area. Then CACREL refers to the probability that a data stream oriented to CIC that meets the sensing service requirements generated by a subset of cs state nodes V′ can reach the sink node.

[0110] According to step (2), it can be known whether the entire wireless sensor network can meet the sensing service requirements can be calculated through the coverage table of the sink node. Based on the definitions of virtual reconstruction nodes and virtual networks, after converting an ordinary WSN into a virtual WSN, the CACREL calculation formula of the network is:

[0111] S min = S square × Areq

[0112]

[0113] (4) Calculate the coverage table of the VRN based on the trusted information coverage model;

[0114] (4.1) Calculate the trusted information coverage in different working states within each grid area;

[0115] (4.1.1) If the grid g contains y three-state sensor nodes, the set of possible working states in this grid area is:

[0116]

[0117] (4.1.2) Enumerate all possible working states, and in each working state, use the sensor nodes in the cs state to perform collaborative information reconstruction on the sensed data of the reconstruction points, and calculate the root mean square error RMSE, as Figure 3 shown.

[0118] (4.1.3) In the trusted information coverage model, for the reconstruction point x i , use the ordinary Kriging interpolation function to calculate the estimated value of the environmental variable of the reconstruction point x i , that is, calculate the estimated value of the environmental variable by taking the weighted average of the measured values of the sensor nodes in the cs state within the reconstruction neighborhood Z(x i ); the interpolation weight coefficient ω i of the sensor nodes within the neighborhood satisfies is the number of sensor nodes s i within the reconstruction neighborhood Z(x i );

[0119] (4.1.4) Combine the ordinary Kriging interpolation function to calculate the root mean square error Φ(x) of the reconstruction point x, and the calculation expression is: where and μ(x) are solved by the following formula;

[0120] (4.1.5) The interpolation weight coefficient λ i obtains a set of optimal solutions through the minimum Kriging variance; introduce the Lagrange multiplier μ(x) to generate a linear Kriging system composed of n + 1 equations with n + 1 unknowns, and solve to obtain the interpolation weight coefficient λ i ;

[0121]

[0122] where, γ(v i , v j ) and γ(v i, x) is obtained by calculating the variogram;

[0123] (4.1.6) Calculate γ(v i , v j ) and γ(v i , x); Select the Gaussian variogram as the variogram of the environmental variable to describe the spatial correlation between the data collected by the sensor node v i ; The formula of the Gaussian variogram is:

[0124]

[0125]

[0126] Where, is the Euclidean distance between the sensor node v i and the reconstruction point x, is the Euclidean distance between the sensor node v i and v j ; C0 and C1 are both constants;

[0127] (4.1.7) According to the definition of the credible information coverage model, if that is, the time-averaged root mean square error is greater than the set coverage threshold, then the grid area in this working state meets the credible information coverage requirement, otherwise it is not covered, and this invalid working state is ignored and the enumeration continues;

[0128] (4.2) Calculate the probability that the working state satisfying the credible information coverage requirement occurs; For any working state The probability that the network operates in this state is calculated by the following formula:

[0129]

[0130] Where, V g is the set of sensor nodes located in the grid g, is the cs state sensor set, is the c state sensor set, is the f state sensor set.

[0131] (4.3) Construct the coverage table C VRN of the VRN; Each a sense (VRN) of the VRN includes two numerical values, as follows:

[0132]

[0133] Where, S g represents the area of the grid, and the area value of the grid is usually CR×CR m 2, but if the field cannot be accurately divided by CR, the area of some grids will be less than CR×CR m 2 .

[0134] Coverage table C of VRN VRN Contains 2 entries as follows:

[0135]

[0136] C VRN [0] = 1 - C VRN [S g - P f (VRN)

[0137] (5) Calculate the reliable coverage table convergence path according to the connection status between each VRN;

[0138] (5.1) Calculate VM; for the network W = {{v sink}} ∪ V, E}, where the communication link e ij ∈E is defined as:

[0139]

[0140] Given two adjacent grids g a and g b , VM ab = k means that there are k links connecting the VRNs in g a and g b . VM ab is calculated by the following formula:

[0141]

[0142] (5.2) VM reflects the reliability of the communication connection between VRNs. The larger the value of VM ab , the stronger the connection between VRN a and VRN b . To make each C VRN converge to the sink node more reliably, we set the connection weight matrix h = 1 / VM to represent the selection priority of the links in VM. The smaller h is, the more likely this path will be selected for coverage table convergence, Figure 4 (b) shows Figure 4 (a) the virtual network structure and weight matrix values corresponding to the network.

[0143] (5.3) Based on the connection weight matrix h, use the dijkstra algorithm to generate the minimum spanning tree, and this tree-like path is the reliable coverage table convergence path RPath;

[0144] (6) Combine the coverage tables of the VRN one by one according to the coverage table aggregation path until all the coverage table information is combined into the VRN sink in the coverage table of the node;

[0145] (6.1) Combined calculation of the coverage table; For two wireless sensor networks W1 and W2, if the two networks only share the same sink node and the rest of the nodes are disjoint, then the coverage table of the network formed by combining W1 and W2 is:

[0146]

[0147] where C1 is the coverage table of W1, C2 is the coverage table of W2, and C1×C2 represents the combined operation of the coverage table; If the coverage table C i has n i coverage records, then C[m] has at most n1×n2 coverage records;

[0148] (6.2) Calculate the reliability evaluation value CACREL of the output network; According to the reliable coverage table aggregation path RPath calculated in (5.3), as Figure 5 shown, starting from the leaf node, perform the coverage table combination operation with its parent node one by one according to (6.1); After all leaf nodes have performed the combination operation, remove all leaf nodes; Continuously perform the combination-removal operation until the coverage tables of all nodes are combined and only the VRN sink node remains in the network; Obtain the final reliability evaluation value CACREL of the entire network according to the CACREL calculation formula;

[0149] It is easy for those skilled in the art to understand that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage, characterized in that Including the following steps: (1) Establish a network model according to the monitoring coverage scenario of the target area; (1.1) Determine the number, location, sensing radius, communication radius, working probability of the sensor sensing module , and the working probability of the sensor communication module . During the operation of the three-state sensor node model, there are the following states: represents the fully operational state, represents the relay state with only communication, represents the completely failed state. The working probabilities of different states are as follows: (1.2) Model the wireless sensor network WSN as an undirected graph W ={{ v 𝑠𝑖𝑛𝑘}∪ V , E}; (1.3) Set an appropriate range according to the spatial correlation of the monitoring variables CR and the root mean square error RMSE , and divide the target area into several grid areas according to the range , and the center of each grid area is the reconstruction point; (2)Construct the coverage table for each sensor node according to the polymorphism of the sensor node, denoted as ; The method for constructing the coverage table in step (2) is as follows: For any multi-state sensor node in the network , its state set is represented as , represents the sensing area of the sensor node in state. Then the possible area values that the sensor node may sense are: Sensor node coverage table is structured as follows: ； (3) Deploy virtual reconstruction nodes in each grid area , construct a virtual network; define the coverage reliability of a wireless sensor network based on trusted information coverage (4) Calculate based on the trusted information coverage model The coverage table; (5) According to the connection status between them, calculate a reliable coverage table aggregation path; (6)Aggregate the paths according to the coverage table, and perform a combination operation on each coverage table of until all the coverage table information is combined into the coverage table of the VRN sink node.

2. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 1, wherein The virtual reconstruction node in step (3) refers to: a virtual sensor node set at the reconstruction point position of each grid area; the virtual reconstruction node is similar to an ordinary sensor, has multiple working states and has a coverage table describing the node's coverage ability , and its working state is jointly determined by all the nodes in the grid, and its coverage table describes the coverage reliability of the grid area where it is located.

3. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 1, wherein, In step (3), the virtual network refers to a virtual wireless sensor network constructed by nodes, denoted as , where represents the virtual sensor node of the grid where the sink node is located, represents the set of communication links between 4. The reliability evaluation method for wireless sensor network coverage based on trusted information coverage according to claim 1, wherein Coverage Reliability Based on Trusted Information Coverage Model The definition means that for a WSN deployed in an area, is the coverage threshold required by the sensing service, indicating the percentage of the minimum coverage area. Then it means that there exists a subset of status nodes , and the probability that the CIC-oriented data flow generated by itself and meeting the sensing service requirements can reach the sink node.

5. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 4, wherein In the step (3), according to step (2), it can be calculated from the coverage table of the sink node whether the entire wireless sensor network can meet the requirements of the sensing service. After converting the ordinary WSN into a virtual WSN, the calculation formula is: 。 6. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 1, wherein The specific steps of step (4) are as follows: (4.1) Calculate the coverage of credible information in different working states within each grid area; (4.2) Calculate the probability of the occurrence of the operating state that meets the trusted information coverage requirement; for any operating state , the probability that the network operates in this state is calculated by the following formula: Among them, is the set of sensor nodes located within the grid , is the set of status sensors, is the set of status sensors, is the set of status sensors; (4.3) Construct the coverage table of the VRN ; For each VRN includes two quantitative values as follows: Among them, represents the area of the grid, and the area value of the grid is usually , but if the field cannot be exactly divided, the area of some grids will be less than ; Coverage table Contains 2 entries as follows: 。 7. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 6, characterized in that, The specific sub-steps of step (4.1) are as follows: (4.1.1)If the grid contains three-state sensor nodes, then the set of possible operating states in this grid area is: (4.1.2)Enumerate all possible working states, and use the sensor nodes in the state to perform collaborative information reconstruction on the sensed data of the reconstruction points, and calculate the root mean square error ; (4.1.3) In the trusted information coverage model, for the reconstruction point , the ordinary Kriging interpolation function is used to calculate the estimated value of the environmental variable at the reconstruction point , that is, the weighted average of the measured values of the sensor nodes in the reconstruction neighborhood in the state is used to calculate the estimated value of the environmental variable; the interpolation weight coefficient of the sensor nodes in the neighborhood satisfies , is the number of sensor nodes in the reconstruction neighborhood ; ​ (4.1.4) Calculate the reconstructed points by combining with the ordinary Kriging interpolation function The root mean square error is calculated by the expression: where and are solved by the following formula; (4.1.5) Interpolation weight coefficient Obtain a set of optimal solutions through the minimum Kriging variance; introduce the Lagrange multiplier Generate a linear Kriging system consisting of equations with unknowns. After solving, the interpolation weight coefficient is obtained; Among them, and are obtained by calculating the variogram. (4.1.6) Calculate the and ; Select the Gaussian variogram as the variogram of the environmental variable to describe the spatial correlation between the data collected by the sensor nodes; The formula of the Gaussian variogram is: Among them, is the Euclidean distance between the sensor node and the reconstruction point ; is the Euclidean distance between the sensor node and ; and are both constants. (4.1.7) According to the definition of the reliable information coverage model, if , that is, the time-averaged root mean square error is greater than the set coverage threshold, then this grid area meets the reliable information coverage requirements in this working state; otherwise, it is not covered, and this invalid working state is ignored and the enumeration continues.

8. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 1, wherein The specific steps of step (5) are as follows: (5.1) Calculate ; for the network W = { { v 𝑠𝑖𝑛𝑘 } ∪ V , E }, where the communication link between nodes is defined as: Given two adjacent grids and , indicating that there is a link with connecting and in ; Calculated by the following formula: (5.2) Reflects The reliability of the communication connection between The larger the value of And The stronger the connection between; In order to make each More reliably converge to Node, we set the connection weight matrix To represent The selection priority of the links in; The smaller, the more likely this path is to be selected for overlay table aggregation; (5.3) Based on the connection weight matrix , use algorithm to generate a minimum spanning tree, and this tree-like path is a reliable coverage table aggregation path .

9. The method for evaluating the coverage reliability of a wireless sensor network based on trusted information coverage according to claim 1, wherein, The specific steps of step (6) are as follows: (6.1)Combined calculation of coverage tables; for two wireless sensor networks , , if the two networks only share the same nodes and the remaining nodes are all disjoint, then the coverage table of the network composed of network and is as follows: Among them is 's coverage table, is 's coverage table, represents the combined operation of the coverage tables; if the coverage table has coverage records, then has at most coverage records; (6.2) Calculate and output the reliability evaluation value of the network ; The reliable coverage table convergence path calculated according to (5.3) , starting from the leaf nodes, perform the coverage table combination operation with its parent nodes one by one according to (6.1); after all leaf nodes have performed the combination operation, remove all leaf nodes; continuously perform the combination-removal operation until the coverage tables of all nodes are combined, and only nodes remain in the network; according to the calculation formula, obtain the final reliability evaluation value of the entire network .

Images