Distributed positioning method based on dual privacy protection strategies

Abstract

The invention discloses a distributed positioning method based on a dual privacy protection strategy, which belongs to the technical field of wireless communication networks, and comprises the following steps: arranging n nodes to be positioned in a convex hull formed by three beacon nodes; presetting a moment set, and randomly setting initial position estimation and relaxation parameters of each node to be positioned; setting a communication radius, a communication radius increment and a generated position updating moment set of the node to be positioned; generating a triangulation measurement set; and obtaining final position estimation of the to-be-positioned node based on a position updating strategy generated based on noise injection and an asynchronous iteration strategy and a distributed iteration positioning method represented by barycentric coordinates. Aiming at a wireless network with time-varying communication time delay, an asynchronous iteration strategy of node intermittent updating is adopted, an information hiding method of noise injection is combined, and the method has the advantages of protecting node position privacy, saving communication overhead and being high in positioning precision.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless communication networks, and in particular to a distributed positioning method based on a dual privacy protection strategy. Background Art

[0002] With the continuous development of wireless network technology, data security and privacy protection issues have gradually become a hot topic of research. Since wireless networks are usually deployed in harsh environments, ignoring privacy protection issues will lead to many malicious attacks on network nodes. For example, some malicious nodes illegally obtain the location information of beacon nodes with known locations, and then attack the beacon nodes to make the positioning function based on the beacon nodes invalid. In addition, malicious nodes can also associate the distance measurements intercepted by eavesdropping with the monitored area to infer the location of other nodes, thereby destroying the entire system. Therefore, how to formulate corresponding security policies to achieve privacy protection of wireless network node location information is an urgent problem to be solved.

[0003] At present, the solutions to the privacy protection problem can be roughly divided into encryption-based methods, state decomposition methods, and differential privacy methods. Among them, the encryption-based method has higher security, but due to its complex authentication process, the amount of calculation is high. The state decomposition method is suitable for the average consistency problem, but not for the distributed positioning problem. The principle of the differential privacy method is to protect privacy by adding structured noise to the location information. This method can improve the computational efficiency, but at the same time it will also reduce the positioning accuracy.

[0004] In addition, there is often a certain delay in the communication between nodes in actual scenarios. Most of the existing research on distributed positioning algorithms adopts a node continuous update strategy, that is, all nodes to be located update their position estimates at each iteration. Under the node continuous update rule, malicious nodes that illegally monitor wireless channels can easily steal the information sent and received by nearby nodes, and infer the exact location of the node, which means that the node continuous update rule is not enough to protect the location privacy of the node. At the same time, updating the location at every moment also increases the node communication overhead, which not only increases the operation and maintenance costs, but also may cause network congestion. Summary of the invention

[0005] The purpose of the present invention is to provide a distributed positioning method based on a dual privacy protection strategy to solve the above problems and achieve high-precision positioning while ensuring the privacy protection of node locations.

[0006] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is as follows: a distributed positioning method based on a dual privacy protection strategy, applied to distributed positioning of a wireless network with time-varying communication delay, comprises the following steps; S1: n nodes to be located are placed in the convex hull formed by three beacon nodes. ~ , where the i-th node to be located is , 1≤i≤n; S2, preset time set ={1,2,…,t,…}, t is the time t, and the initial position estimation and relaxation parameters of each node to be located are randomly set, where The initial position of is estimated to be , the relaxation parameter is γ; S3, Settings Communication radius , Communication radius increment ,exist In the above example, select K moments from the beginning to the end and mark them as ~ ,generate The location update time collection ,in is the kth update time, 1≤k≤K; S4, for Generate a triangulation set that satisfies condition 1 , including steps S41 to S43; S41, will exist The adjacent nodes under the node set constitute an adjacent node set, and three adjacent nodes are randomly selected to form a triangle. The first preset condition is: Located within the triangle; S42, test Whether condition 1 is met, if so, execute step S43, otherwise Increase , repeat S41 until condition 1 is met; S43, mark the three adjacent nodes that meet condition 1 as , , , the three constitute the triangulation set ; S5, based on , noise injection and asynchronous iterative strategy generation The location update strategy comprises steps S51 to S53; S51, obtain The initial position estimate and position update time set of each node to be located; S52, in At each moment, a noise term is generated for all nodes to be located. The noise term at time t is ; S53, using the following formula as the update strategy, generates The position estimate at time t+1 ; , , In the formula, for The nodes in , , , , They are The position estimate at time t, the position estimate after injecting noise, For time t from arrive The time lag, for time Position estimation after noise injection, for Relative to The centroid coordinates of S6, obtaining a final position estimate of the node to be located based on a distributed iterative positioning method.

[0007] As a preferred embodiment: testing in step S42 Whether the condition 1 is met, the To Left Test method is used to judge Whether it is located in the convex hull formed by three adjacent nodes.

[0008] As a preference: calculate a certain moment, Relative to , , The centroid coordinates of , , The methods include Sa1~Sa3; Sa1, measure the moment, and , , Distance , , , and Distance , and Distance , and Distance ; Sa2, the area of ​​a triangle based on the Cayley-Menger determinant , , , ,in for , , The area of ​​the triangle formed, , , The area of ​​the triangle formed, , , The area of ​​the triangle formed, , , The area of ​​the triangle formed; Sa3, calculation Relative to , , The centroid coordinates of , , ,in, , , .

[0009] Preferably, the distance measurement method includes laser radar ranging, RSSI ranging, TOA ranging, TDOA ranging and AOA ranging.

[0010] As a preference: The noise term at time t is According to the following formula: , , In the formula, is a constant and , for The noise value at time t is, for The noise value at time t is, is the preset convergence time.

[0011] As a preferred embodiment: step S6 is specifically as follows; S61, iteratively calculating a final position estimate of the node to be located according to the following matrix equation; , in, , , They are time t, t+1, The coordinate vector of the beacon node into which the noise is injected, , ~ are the coordinates of the three beacon nodes injecting noise at time t, and T is the matrix transpose; , , They are time t, t+1, The corresponding position estimate of the node to be located, , is the n-order identity matrix; is the indicator function, is the Kronecker product operation; , is the noise matrix injected into the node to be located at time t; The time lag is The sub-random matrix of the beacon node set, The time lag is The sub-random matrix of the set of nodes to be located, , is the preset upper bound of communication delay; S62, iteratively solve the formula in S61 to obtain a position solution of the node to be located based on the beacon node with a known position as follows; , Where F and H are the sub-random matrices corresponding to the beacon node set and the node set to be located, respectively. The coordinate vector of the beacon node that injects noise at the initial moment.

[0012] In the present invention, regarding steps S61 and S62, after obtaining the formula of S61, in S62, the formula of S61 is iteratively converged and solved by utilizing the composite relationship of directed edge sets in graph theory and the sub-random matrix product property and the matrix infinite norm property in matrix theory, so as to obtain a position solution represented by the beacon node set with known positions of the node set to be located.

[0013] Compared with the prior art, the advantages of the present invention are: mainly for wireless networks with time-varying communication delays, a distributed positioning method based on asynchronous iteration and noise injection dual privacy protection strategies is proposed, and the following improvements are mainly made on the basis of the existing distributed positioning method.

[0014] (1) The present invention adopts an asynchronous iterative strategy of intermittent node updates: the present invention randomly and independently selects a location update time for each node to be located and generates a respective location update time set, such as The location update time collection , Each update moment is from ={1,2,…,t,…} is randomly selected, so the update time of each node to be located is different. In this strategy, each node to be located does not update its position estimate at the same time, but relatively independently. Updating its position estimate not only greatly reduces the possibility that the information sent and received by the node is stolen by malicious nodes at the same time, but also reduces the communication traffic between nodes, avoiding network congestion or even packet loss due to node resource exhaustion.

[0015] (2) This paper considers an information hiding strategy based on noise injection, which injects a pre-designed noise term into the position estimate transmitted between nodes, thereby achieving accurate positioning of nodes while protecting the privacy of location information. However, unlike the structured noise injected in the differential privacy method, which reduces the positioning accuracy, the noise term in this paper By design, the added noise will eventually cancel out completely and therefore has no impact on positioning accuracy.

[0016] (3) In summary, the present invention adopts an asynchronous iterative strategy of intermittent node updates and combines it with an information hiding method of noise injection for wireless networks with time-varying communication delays. It has the advantages of protecting node location privacy, saving communication overhead, and high positioning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a diagram of the experimental platform setting in Example 2; Figure 2 This is the software interface display diagram of NOKOV motion capture system; Figure 3 The position estimation trajectory generated by method 1 in embodiment 2; Figure 4 The position estimation trajectory generated by method 2 in embodiment 2; Figure 5 is the position estimation error of method 1 in embodiment 2; Figure 6 is the position estimation error of method 2 in embodiment 2; Figure 7 The position estimation trajectory of the unmanned vehicle using the asynchronous iterative strategy in Example 3; Figure 8 The position estimation trajectory of the unmanned vehicle using the synchronous iteration strategy in Example 3; Fig. 9 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0018] The present invention will be further described below in conjunction with embodiments and drawings.

[0019] Example 1: See Fig. 9 ,A distributed positioning method based on a dual privacy protection strategy is applied to ,distributed positioning in wireless networks with time-varying communication delay, including the following steps; S1, deploy n nodes to be located in the convex hull composed of three beacon nodes ~ , where the i-th node to be located is , 1≤i≤n; S2, preset time set ={1,2,…,t,…}, t is the time t, and the initial position estimation and relaxation parameters of each node to be located are randomly set, where The initial position of is estimated to be , the relaxation parameter is γ; S3, Settings Communication radius , Communication radius increment ,exist In the above example, select K moments from the beginning to the end and mark them as ~ ,generate The location update time collection ,in is the kth update time, 1≤k≤K; S4, for Generate a triangulation set that satisfies condition 1 , including steps S41 to S43; S41, will exist The adjacent nodes under the node set constitute an adjacent node set, and three adjacent nodes are randomly selected to form a triangle. The first preset condition is: Located within the triangle; S42, test Whether condition 1 is met, if so, execute step S43, otherwise Increase , repeat S41 until condition 1 is met; S43, mark the three adjacent nodes that meet condition 1 as , , , the three constitute the triangulation set ; S5, based on , noise injection and asynchronous iterative strategy generation The location update strategy comprises steps S51 to S53; S51, obtain The initial position estimate and position update time set of each node to be located; S52, in At each moment, a noise term is generated for all nodes to be located. The noise term at time t is ; S53, using the following formula as the update strategy, generates The position estimate at time t+1 ; , , In the formula, for The nodes in , , , , They are The position estimate at time t, the position estimate after injecting noise, For time t from arrive The time lag, for time Position estimation after noise injection, for Relative to The centroid coordinates of S6, obtaining a final position estimate of the node to be located based on a distributed iterative positioning method.

[0020] In this embodiment: in step S42, test Whether the condition 1 is met, the To Left Test method is used to judge Whether it is located in the convex hull formed by three adjacent nodes.

[0021] In this embodiment, a certain moment is calculated, Relative to , , The centroid coordinates of , , The methods include Sa1~Sa3; Sa1, measure the moment, and , , Distance , , , and Distance , and Distance , and Distance ; Sa2, the area of ​​a triangle based on the Cayley-Menger determinant , , , ,in for , , The area of ​​the triangle formed, , , The area of ​​the triangle formed, , , The area of ​​the triangle formed, , , The area of ​​the triangle formed; Sa3, calculation Relative to , , The centroid coordinates of , , ,in, , , .

[0022] The distance measurement methods include lidar ranging, RSSI ranging, TOA ranging, TDOA ranging and AOA ranging.

[0023] The noise term at time t is According to the following formula: , , In the formula, is a constant and , for The noise value at time t is, for The noise value at time t is, is the preset convergence time.

[0024] Step S6 specifically comprises: S61, iteratively calculating a final position estimate of the node to be located according to the following matrix equation; , in, , , are time t, t+1, The coordinate vector of the beacon node into which the noise is injected, , ~ are the coordinates of the three beacon nodes injecting noise at time t, and T is the matrix transpose; , , are time t, t+1, The corresponding position estimate of the node to be located, , is the n-order identity matrix; is the indicator function, is the Kronecker product operation; , is the noise matrix injected into the node to be located at time t; The time lag is The sub-random matrix of the beacon node set, The time lag is The sub-random matrix of the set of nodes to be located, , is the preset upper bound of communication delay; S62, iteratively solve the formula in S61 to obtain a position solution of the node to be located based on the beacon node with a known position as follows; , Where F and H are the sub-random matrices corresponding to the beacon node set and the node set to be located, respectively. The coordinate vector of the beacon node that injects noise at the initial moment.

[0025] Regarding Sa2, we give an example of obtaining the area of ​​a triangle based on the Cayley-Menger determinant. Suppose we calculate , the Cayley-Menger determinant is as follows: , According to this formula, we can get The square root of .

[0026] Example 2: See Figures 1 to 6 ,In order to test the performance of the present invention, an ,experimental platform consisting of 9 unmanned vehicles was established to ,simulate a wireless network. The experimental settings are as follows. Figure 1As shown in the figure, there are 9 unmanned vehicles, including 5 Tianbot unmanned vehicles, 2 Turtlebot2 unmanned vehicles and 2 Robomaster Ep unmanned vehicles, marked as R1~R9 respectively. Each unmanned vehicle is equipped with a laser radar for ranging, and multiple cameras are arranged around the unmanned vehicle to collect video data and monitor in real time through the NOKOV motion capture system. Figure 2 The software interface of the NOKOV motion capture system is displayed. Figure 2 The 9 light spots in the middle are unmanned vehicles, and the cameras are placed around them. This embodiment uses the following two methods to estimate the position of the unmanned vehicle and generate a position estimation trajectory; Method 1: The method of the present invention, the position estimation is the asynchronous iterative strategy described in the present invention.

[0027] Method 2: The asynchronous iteration strategy of the present invention is replaced by a synchronous iteration strategy, that is, at each time step, each node to be located performs a position estimation.

[0028] Based on method 1 and method 2, distributed positioning is performed to generate position estimation trajectories as follows: Figure 3 , Figure 4 shown. Figure 3 , Figure 4 In the figure, the green diamond, blue star and red box represent the initial position estimate, final position estimate and actual position of the unmanned vehicle respectively. The line connecting the green diamond and blue star is the position estimate trajectory of the corresponding unmanned vehicle. Figure 3 , Figure 4 It can be seen that the position estimation of the two methods eventually converges to its actual position, that is, accurate positioning is achieved. However, the position estimation trajectory of the present invention is slightly jagged, while the position estimation trajectory of method 2 is relatively smooth. This is because the intermittent update of the present invention adopts an asynchronous iterative strategy, which has the dual advantages of high privacy protection performance and low network communication overhead compared with the synchronous continuous update strategy.

[0029] In addition, the position estimation errors of method 1 and method 2 are also counted. Figure 5 , Figure 6 As shown, from Figure 5 , Figure 6 It can be seen that the convergence speed of the node asynchronous iteration strategy described in method one is slightly slower than that of the node synchronous iteration strategy described in method two, but the position estimation error of the unmanned vehicle to be located under the two update strategies eventually approaches 0; and when the asynchronous iteration strategy is adopted, the average position estimation error has been reduced by nearly 80% at t=20. Considering the positioning performance and communication overhead, the node intermittent update strategy based on asynchronous iteration proposed in the present invention has certain advantages. Figure 3~Figure 6From the experimental results in , we can see that the position estimates of the six unmanned vehicles to be located finally converge to their precise positions, that is, the distributed positioning method based on the dual privacy protection strategy proposed in the present invention achieves precise positioning.

[0030] Example 3: See Figure 7 , Figure 8 , consider the case where malicious nodes can steal the information received and sent by network nodes at the same time, and compare the privacy protection performance of asynchronous iteration strategy and synchronous iteration strategy. Assume that R6 is captured by the attacker and becomes a malicious node, whose purpose is to infer the exact location of R5. At each iteration time t, R6 estimates the location of R5, and its estimated value is recorded as The estimated position of R5 itself is recorded as The actual position of the unmanned vehicle 5 is recorded as . Get the next Figure 7 , Figure 8 .

[0031] from Figure 8 It can be seen that under the node synchronization iteration strategy, and When t→∞, they gradually reach a consensus, that is, R6 successfully infers the exact location of R5, which means that the location privacy information of R5 is leaked and the privacy protection function is not achieved. Figure 7 As can be seen from the table, under the node asynchronous iteration strategy, R6 cannot pass The exact location of R5 is inferred, thus achieving privacy protection function.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A distributed positioning method based on a dual privacy protection strategy, applied to distributed positioning in wireless networks with time-varying communication delays, characterized in that: The steps include: S1: n nodes to be located are placed in the convex hull formed by three beacon nodes. ~ , where the i-th node to be located is , 1≤i≤n; S2, preset time set ={1,2,…,t,…}, t is the time t, and the initial position estimation and relaxation parameters of each node to be located are randomly set, where The initial position of is estimated to be , the relaxation parameter is γ; S3, Settings Communication radius , Communication radius increment ,exist In the above example, select K moments from the beginning to the end and mark them as ~ ,generate The location update time collection ,in is the kth update time, 1≤k≤K; S4, for Generate a triangulation set that satisfies condition 1 , including steps S41 to S43; S41, will exist The adjacent nodes under the node set constitute an adjacent node set, and three adjacent nodes are randomly selected to form a triangle. The first preset condition is: Located within the triangle; S42, test Whether condition 1 is met, if so, execute step S43, otherwise Increase , repeat S41 until condition 1 is met; S43, mark the three adjacent nodes that meet condition 1 as , , , the three constitute the triangulation set ; S5, based on , noise injection and asynchronous iterative strategy generation The location update strategy comprises steps S51 to S53; S51, obtain The initial position estimate and position update time set of each node to be located; S52, in At each moment, a noise term is generated for all nodes to be located. The noise term at time t is ; S53, using the following formula as the update strategy, generates The position estimate at time t+1 ; ， ， In the formula, for The nodes in , , , , They are The position estimate at time t, the position estimate after injecting noise, For time t from arrive The time lag, for time Position estimation after noise injection, for Relative to The centroid coordinates of S6, obtaining a final position estimate of the node to be located based on a distributed iterative positioning method.

2. The distributed positioning method based on the dual privacy protection strategy according to claim 1, characterized in that: Test in step S42 Whether the condition 1 is met, the To Left Test method is used to judge Whether it is located in the convex hull formed by three adjacent nodes.

3. The distributed positioning method based on the dual privacy protection strategy according to claim 1, characterized in that: Calculate a moment, Relative to , , The centroid coordinates of , , The methods include Sa1~Sa3; Sa1, measure the moment, and , , Distance , , , and Distance , and Distance , and Distance ; Sa2, the area of ​​a triangle based on the Cayley-Menger determinant , , , ,in for , , The area of ​​the triangle formed, , , The area of ​​the triangle formed, , , The area of ​​the triangle formed, , , The area of ​​the triangle formed; Sa3, calculation Relative to , , The centroid coordinates of , , ,in, , , .

4. The distributed positioning method based on the dual privacy protection strategy according to claim 3 is characterized in that: The distance measurement methods include lidar ranging, RSSI ranging, TOA ranging, TDOA ranging and AOA ranging.

5. The distributed positioning method based on a dual privacy protection strategy according to claim 1, characterized in that: The noise term at time t is According to the following formula: ， ， In the formula, is a constant and , for The noise value at time t is, for The noise value at time t is, is the preset convergence time.

6. The distributed positioning method based on a dual privacy protection strategy according to claim 1, characterized in that: Step S6 specifically comprises: S61, iteratively calculating a final position estimate of the node to be located according to the following matrix equation; ， in, , , are time t, t+1, The coordinate vector of the beacon node into which the noise is injected, , ~ are the coordinates of the three beacon nodes injecting noise at time t, and T is the matrix transpose; , , They are time t, t+1, The corresponding position estimate of the node to be located, , is the n-order identity matrix; is the indicator function, is the Kronecker product operation; , is the noise matrix injected into the node to be located at time t; The time lag is The sub-random matrix of the beacon node set, The time lag is The sub-random matrix of the set of nodes to be located, , is the preset upper bound of communication delay; S62, iteratively solve the formula in S61 to obtain a position solution of the node to be located based on the beacon node with a known position as follows; ， Where F and H are the sub-random matrices corresponding to the beacon node set and the node set to be located, respectively. The coordinate vector of the beacon node that injects noise at the initial moment.

Images