Multi-AUV (Autonomous Underwater Vehicle) cooperative low-delay underwater acoustic network source node position privacy protection method

Abstract

The invention provides a multi-AUV (Autonomous Underwater Vehicle) cooperative low-delay underwater acoustic network source node position privacy protection method, which belongs to the technical field of underwater acoustic network defense and comprises the following steps: dividing a three-dimensional underwater acoustic network into a plurality of three-dimensional sub-regions by adopting a k-dimensional tree method, the underwater acoustic network comprises a plurality of sensor nodes, a plurality of autonomous underwater vehicles and a base station node on the water surface, and any underwater sensor node is a source node; determining hierarchical information corresponding to each sub-region, a neighbor table corresponding to each sensor node and a living region corresponding to each autonomous underwater vehicle; data packets are transmitted to the base station node from the source node through the sensor nodes and the autonomous underwater vehicles through transmission of the multi-constraint routing node and cooperative transmission of the multiple autonomous underwater vehicles, and the living area is the area where the AUV collects the data packets; according to the invention, more dispersed and diversified transmission paths are realized, and the energy consumption and time delay of the underwater acoustic network are greatly reduced while the location privacy of the source node is protected.

Description

Technical Field

[0001] This application belongs to the technical field of underwater acoustic network defense, and particularly relates to a method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multi-AUV cooperation. Background Art

[0002] In wireless networks such as wireless sensor networks (WSNs) and underwater acoustic networks (hereinafter referred to as UANs, Underwater Acoustic Networks), the source node carries important data. Once the source node is captured, it may cause serious consequences such as physical attacks on the source node or leakage of important information.

[0003] The goal of source node location privacy protection is that it is difficult for attackers to accurately determine the location of the source node even when implementing passive attacks. The source node location privacy in UANs means that while ensuring the normal transmission of data packets from the source node to the base station (Sink) node, it is necessary to prevent attackers from finding the source node location as much as possible or reduce the probability of attackers finding the source node location.

[0004] Existing technologies have proposed some methods for protecting the location privacy of source nodes through the cooperation of autonomous underwater vehicles (hereinafter referred to as AUVs, Autonomous Underwater Vehicles) in UANs. Although these methods have achieved source node location privacy protection, using AUVs for transmission will cause long delays. In water, AUVs usually move at a speed of about 8 m / s, which is much lower than the speed of sound in water (1500 m / s). This difference leads to long delay problems when using AUVs to collect and transmit data. The multi-path method will reduce network performance, including increased energy consumption and increased end-to-end delay, and the centralized multiple transmission paths will reduce the privacy security of the source node location.

[0005] The existing underwater acoustic networks have problems of high energy consumption and long delay when protecting the location privacy of source nodes. Summary of the Invention

[0006] The embodiments of this application provide a method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multi-AUV cooperation, which can solve the problems of high energy consumption and long delay when the underwater acoustic network protects the location privacy of source nodes.

[0007] In a first aspect, the embodiments of this application provide a method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multi-AUV cooperation, including:

[0008] The method of using a k - dimensional tree divides a three - dimensional underwater acoustic network into multiple three - dimensional sub - regions. The underwater acoustic network includes multiple underwater sensor nodes, multiple autonomous underwater vehicles, and a base station node on the water surface. Any underwater sensor node is a source node;

[0009] Based on each sub - region and each sensor node, determine the hierarchical information corresponding to each sub - region, the neighbor table corresponding to each sensor node, and the residence area corresponding to each autonomous underwater vehicle;

[0010] Based on the hierarchical information corresponding to each sub - region, the neighbor table corresponding to each sensor node, and the residence area corresponding to each autonomous underwater vehicle, through multi - constraint routing node transmission and multi - autonomous underwater vehicle cooperative transmission, transmit the data packet from the source node through each sensor node and each autonomous underwater vehicle to the base station node. The residence area is the area where the autonomous underwater vehicle collects the data packet.

[0011] The beneficial effects of the embodiment of this application compared with the prior art are:

[0012] The method for protecting the location privacy of the source node in a low - latency underwater acoustic network with multi - AUV cooperation in this application divides a three - dimensional underwater acoustic network into multiple three - dimensional sub - regions by using the method of a k - dimensional tree. The underwater acoustic network includes multiple underwater sensor nodes, multiple autonomous underwater vehicles, and a base station node on the water surface. Any underwater sensor node is a source node; based on each sub - region and each sensor node, determine the hierarchical information corresponding to each sub - region, the neighbor table corresponding to each sensor node, and the residence area corresponding to each autonomous underwater vehicle; based on the hierarchical information corresponding to each sub - region, the neighbor table corresponding to each sensor node, and the residence area corresponding to each autonomous underwater vehicle, through multi - constraint routing node transmission and multi - autonomous underwater vehicle cooperative transmission, transmit the data packet from the source node through each sensor node and each autonomous underwater vehicle to the base station node. The residence area is the area where the autonomous underwater vehicle collects the data packet. Compared with the prior art, because multi - constraint sensor nodes are used for routing node transmission, while ensuring the location privacy security of the source node, the network energy consumption and latency are reduced; because each autonomous underwater vehicle is assigned to the residence area and scheduled, the latency of the AUV is reduced and the scheduling efficiency of the AUV is improved; when multi - constraint routing node transmission and multi - autonomous underwater vehicle cooperative transmission are combined, a more dispersed and diverse transmission path is realized, expanding the search range of the attacker and eliminating the latency caused by the data packet waiting for the AUV to collect and the AUV cruising; thus, when protecting the location privacy of the source node, the energy consumption and latency of the underwater acoustic network are greatly reduced. Description of the Drawings

[0013] To more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the accompanying drawings required for use in the embodiments or the description of the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0014] Figure 1 is a schematic flowchart of a method for protecting the location privacy of the source node in a low-latency underwater acoustic network with multi-AUV collaboration in the present application;

[0015] Figure 2 is a schematic diagram of a UANs source location privacy protection model provided by an embodiment of the present application;

[0016] Figure 3 is a schematic flowchart of step S31 in an embodiment of the present application for determining the target area of a data packet based on the hierarchical information corresponding to each sub-region and the grey correlation degree value corresponding to each sub-region;

[0017] Figure 4 is a schematic flowchart of step S322 in an embodiment of the present application for sending a data packet to the best next-hop node through a multi-constrained routing node transmission model based on the neighbor table corresponding to each sensor node;

[0018] Figure 5 is a schematic flowchart of step S3222 in an embodiment of the present application for determining the cost function value corresponding to each candidate node through a multi-constrained routing node transmission model;

[0019] Figure 6 is a schematic diagram of the forwarding angle and the forwarding angle constraint value in the present application;

[0020] Figure 7 is a schematic diagram of the influence of the weight coefficient η of the objective function on the safety period in an embodiment of the present application;

[0021] Figure 8 is a schematic diagram of the influence of the weight coefficient η of the objective function on the end-to-end delay in the present application;

[0022] Figure 9 is a schematic diagram of the influence of the communication radius on the safety period in an embodiment of the present application;

[0023] Figure 10 is a schematic diagram of the influence of the communication radius on the end-to-end delay in an embodiment of the present application;

[0024] Figure 11 is a schematic diagram of the influence of the communication radius on the average node energy consumption in an embodiment of the present application;

[0025] Figure 12It is a schematic diagram showing the influence of communication radius on the average AUV energy consumption provided by an embodiment of the present application;

[0026] Figure 13 It is a schematic diagram comparing the number of nodes and the safety period of five protection methods of the present application;

[0027] Figure 14 It is a schematic diagram comparing the depth and the safety period of five protection methods of the present application;

[0028] Figure 15 It is a schematic diagram comparing the network side length and the safety period of five protection methods of the present application;

[0029] Figure 16 It is a schematic diagram comparing the number of nodes and the end-to-end delay of five methods of the present application;

[0030] Figure 17 It is a schematic diagram comparing the depth and the end-to-end delay of five methods of the present application;

[0031] Figure 18 It is a schematic diagram comparing the network side length and the end-to-end delay of five methods of the present application;

[0032] Figure 19 It is a schematic diagram comparing the number of nodes and the average node energy consumption of four methods of the present application;

[0033] Figure 20 It is a schematic diagram comparing the depth and the average node energy consumption of four methods of the present application;

[0034] Figure 21 It is a schematic diagram comparing the network side length and the average node energy consumption of four methods of the present application;

[0035] Figure 22 It is a schematic diagram comparing the number of nodes and the average AUV energy consumption of five methods of the present application;

[0036] Figure 23 It is a schematic diagram comparing the depth and the average AUV energy consumption of five methods of the present application;

[0037] Figure 24 It is a schematic diagram comparing the network side length of nodes and the average AUV energy consumption of five methods of the present application. Detailed implementation manners

[0038] In the following description, specific details such as specific system architectures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0039] The goal of source node location privacy protection is that it is difficult for an attacker to accurately determine the location of the source node even when conducting a passive attack. That is, the probability that the source node inferred by the attacker is the same as the real source node is zero, which can be expressed as:

[0040]

[0041] Wherein, represents the location of the source node inferred by the attacker, L s represents the real location of the source node, and P() represents probability.

[0042] Compared with ordinary sensor nodes, AUVs have autonomy, flexibility, and adaptability to complex environments in task execution. Therefore, AUVs are indispensable in complex underwater scenarios.

[0043] To solve the problems of the prior art, the method for protecting the location privacy of the source node in a low-latency underwater acoustic network with multi-AUV cooperation in the present application uses the k-dimensional tree method to divide the three-dimensional underwater acoustic network into multiple three-dimensional sub-regions. The underwater acoustic network includes multiple underwater sensor nodes, multiple autonomous underwater vehicles, and a base station node on the water surface. Any underwater sensor node is the source node; based on each sub-region and each sensor node, the hierarchical information corresponding to each sub-region, the neighbor table corresponding to each sensor node, and the residence area corresponding to each autonomous underwater vehicle are determined; based on the hierarchical information corresponding to each sub-region, the neighbor table corresponding to each sensor node, and the residence area corresponding to each autonomous underwater vehicle, through multi-constrained routing node transmission and multi-autonomous underwater vehicle cooperative transmission, the data packet is transmitted from the source node through each sensor node and each autonomous underwater vehicle to the base station node. The residence area is the area where the autonomous underwater vehicle collects data packets; compared with the prior art, due to using sensor nodes with multiple constraints for routing node transmission, while ensuring the privacy security of the source node location, the network energy consumption and latency are reduced; due to allocating each autonomous underwater vehicle to the residence area and performing scheduling, the latency of the AUV is reduced and the scheduling efficiency of the AUV is improved; when combining multi-constrained routing node transmission and multi-autonomous underwater vehicle cooperative transmission, a more dispersed and diverse transmission path is achieved, expanding the search range of the attacker and eliminating the latency caused by the data packet waiting for the AUV to collect and the AUV cruising; thus, when protecting the location privacy of the source node, the energy consumption and latency of the underwater acoustic network are greatly reduced.

[0044] It should be noted that the method for protecting the source node location privacy of this application is also called the low-delay underwater acoustic network source node location privacy protection method based on multi-AUV collaboration (A Low-Delay Source-Location-Privacy Protection Protocol with Multi-AUV Collaboration for Underwater Acoustic Networks, abbreviated as LDSLP-MA).

[0045] The technical solution of this application will be described below through specific embodiments.

[0046] In the first aspect, as Figure 1 shown, this application provides a method for protecting the source node location privacy of a low-delay underwater acoustic network with multi-AUV collaboration, including:

[0047] S1, using the k-d tree method to divide the three-dimensional underwater acoustic network into multiple three-dimensional sub-regions. The underwater acoustic network includes multiple underwater sensor nodes, multiple autonomous underwater vehicles, and a base station node on the water surface. Any underwater sensor node is the source node.

[0048] In this embodiment, as Figure 2 shown, this embodiment combines the UANs model with the "panda-hunter" model to form a UANs source location privacy protection model. Sensor nodes and AUVs are randomly deployed in the three-dimensional UANs, and a base station (Sink) node is deployed on the water surface. When any sensor node senses a target (panda), it becomes the source node. Even if the source node changes with the change of the target location, there is only one source node in the underwater acoustic network. The data collected by the source node is transmitted from bottom to top to the base station node on the water surface. Using the patient attacker model, the attacker around the base station node analyzes the source of the data packet through eavesdropping and backtracking attacks and gradually moves towards the source node.

[0049] In one embodiment, using the k-d tree method to divide the three-dimensional underwater acoustic network into multiple three-dimensional sub-regions includes:

[0050] S11, establishing a three-dimensional coordinate system with the base station node as the origin and obtaining the three-dimensional coordinates of multiple data points.

[0051] In this embodiment, a three-dimensional coordinate system is established with the base station node on the water surface as the origin, and the three-dimensional coordinates (x i , y i , z i ) of multiple data points are obtained to form a three-dimensional coordinate data set corresponding to the number of data points.

[0052] S12. Based on the coordinate of any data point in one dimension, the average coordinate of all data points in one dimension, and the number of data points, confirm the variance value of one dimension.

[0053] In this embodiment, based on the coordinate of any data point in one dimension, the average coordinate of all data points in one dimension, and the number of data points, confirm the variance value of one dimension through the variance calculation formula; thus, obtain the variance values of three dimensions.

[0054] In one embodiment, the variance calculation formula is:

[0055]

[0056] where d i is the coordinate of the i-th data point in a certain dimension of the three-dimensional coordinate dataset, is the average value of the coordinates of all data points on the same dimension, MSE is the variance value, and a is the number of data points.

[0057] S13. Take the dimension corresponding to the maximum variance value as the splitting axis, and determine the median of the coordinates of all data points on the splitting axis as the splitting point.

[0058] S14. Based on the dimension where the splitting axis is located, divide the data points smaller than the median of the coordinates into the left branch, and divide the data points larger than the median of the coordinates into the right branch.

[0059] S14. Recursively construct the left branch and the right branch with the three dimensions as the splitting axes in turn until the underwater acoustic network is divided into multiple three-dimensional sub-regions.

[0060] It can be understood that according to the coordinate value of the splitting point in a certain dimension, the three-dimensional space is divided into a left sub-region and a right sub-region. The left sub-region contains all the data points in the left branch, and the right sub-region contains all the data points in the right branch. For the obtained left sub-region and right sub-region, recursively execute the above splitting steps respectively. In each recursive splitting, select a new splitting axis, determine a new splitting point and new left and right branches, and further divide the sub-region into smaller sub-regions. The recursive splitting operation continues until each data point is assigned to a single region and the sub-regions are no longer divided. Finally, the entire three-dimensional UANs is irregularly divided into multiple well-defined sub-regions.

[0061] In this embodiment, when the splitting axis is set as the x-axis, the splitting axis L = 1; when the splitting axis is the y-axis, the splitting axis L = 2; when the splitting axis is the z-axis, the splitting axis L = 3; update the splitting axis according to L = (L % 3)+1 (% is the remainder operation), which is convenient to divide the UANs into multiple sub-regions in an irregular way, improves the irregularity of the sub-regions, and increases the difficulty for attackers to obtain the location of the source node.

[0062] S2. Based on each sub-region and each sensor node, determine the hierarchical information corresponding to each sub-region, the neighbor table corresponding to each sensor node, and the dwelling area corresponding to each autonomous underwater vehicle.

[0063] In this embodiment, the hierarchical information of the sub-region where the base station node is located is set to "1". In the direction from the water surface to the underwater, the hierarchical information increases sequentially until the bottom layer of the UANs. The level of each sub-region is determined according to the minimum number of sub-regions required to pass from the current sub-region to the sub-region where the base station node is located. The level includes the current sub-region and the sub-region where the base station node is located. The neighbor table of the sensor node includes information such as the node ID, remaining energy, and three-dimensional coordinates corresponding to the sensor node. The base station node broadcasts hello packets periodically, and other nodes flood the hello packets, so that each node that can be connected to the base station node can obtain the information of its neighbor nodes through the received hello packets. During subsequent data transmission, each node can update its neighbor table by listening to the data packets sent by neighbor nodes.

[0064] In one embodiment, the dwelling area is the area where the autonomous underwater vehicle collects data packets. To protect the location privacy of the source node and reduce the end-to-end delay, the levels of all dwelling areas are the same and the level of the AUV dwelling area is less than the level L of the sub-region where the source node is located. source . At L source ≤ 3, the AUV does not perform data packet transmission. The calculation formula for the set B of the dwelling areas corresponding to each AUV is:

[0065]

[0066] where B is the set of AUV dwelling areas, b is the sub-region, L(b) is the level of the sub-region b, and L source is the level of the sub-region where the source node is located. The calculation formula for the set of dwelling areas ensures that the routing path of the data packet from the source node to the base station node must pass through the specified dwelling area, which is beneficial for the AUV to collect data. In addition, it also ensures that the dwelling area is kept at a certain distance from the sub-region containing the source node, thus avoiding revealing the location of the source node.

[0067] S3. Based on the hierarchical information corresponding to each sub-region, the neighbor table corresponding to each sensor node, and the dwelling area corresponding to each autonomous underwater vehicle, through multi-constraint routing node transmission and multi-autonomous underwater vehicle cooperative transmission, transmit the data packet from the source node through each sensor node and each autonomous underwater vehicle to the base station node.

[0068] In this embodiment, when combining multi-constrained routing node transmission and multi-autonomous underwater vehicle (AUV) cooperative transmission, a more decentralized and diversified transmission path is achieved, expanding the attacker's search range and eliminating the delay caused by data packets waiting for AUV collection and AUV cruising. Thus, when protecting the location privacy of the source node, the energy consumption and delay of the underwater acoustic network are greatly reduced.

[0069] In one embodiment, the target area is the destination for the AUV to transmit data packets. Step S3, based on the hierarchical information corresponding to each sub-area, the neighbor table corresponding to each sensor node, and the residence area corresponding to each AUV, transmits the data packet from the source node through each sensor node and each AUV to the base station node through multi-constrained routing node transmission and multi-AUV cooperative transmission, including:

[0070] S31. Determine the target area of the data packet based on the hierarchical information corresponding to each sub-area and the grey relational grade value corresponding to each sub-area.

[0071] S32. Based on the neighbor table corresponding to each sensor node and the residence area corresponding to each AUV, determine whether any AUV obtains the data packet, and transmit the data packet from the source node through each sensor node and each AUV to the base station node through multi-constrained routing node transmission and multi-AUV cooperative transmission.

[0072] In this embodiment, the target area of the data packet is determined by the hierarchical information corresponding to each sub-area and the grey relational grade value corresponding to each sub-area, which improves the diversity of the path selected by the AUV and reduces the long delay. Then, according to whether the AUV obtains the data packet during multi-constrained routing node transmission, it is decided whether the AUV at the level of the residence area performs cooperative transmission. If the data packet is not obtained, it is transmitted through the multi-constrained routing node, thus combining the data packet jump transmission between sensor nodes and AUV cooperative transmission, and further transmitting the data packet from the source node through each sensor node and each AUV to the base station node, improving the path diversity and reducing the long delay.

[0073] In one embodiment, as Figure 3 shown, step S31 of determining the target area of the data packet based on the hierarchical information corresponding to each sub-area and the grey relational grade value corresponding to each sub-area includes:

[0074] S311. Determine the weight vector corresponding to each evaluation index of the sub-area by means of analytic hierarchy process.

[0075] In this embodiment, the evaluation objects of the target area are all sub-areas other than the sub-area where the base station node is located and with a smaller level than the residential area. The evaluation indicators include the sub-area level L(b), the number of times N that the specified sub-area is the target area, the minimum distance min(d s ) between the sub-area and the base station node, the maximum distance max(d s ) between the sub-area and the base station node, the minimum distance min(d source ) between the sub-area and the source node, and the maximum distance max(d source ) between the sub-area and the source node.

[0076] In this embodiment, the judgment matrix C is determined according to the importance degree of each evaluation indicator for the evaluation object, then the judgment matrix C is normalized, and then the sum is calculated to obtain the weight vector.

[0077] In one embodiment, the importance degree of each evaluation indicator for the evaluation object is shown in the following table.

[0078]

[0079] In one embodiment, the judgment matrix C is as follows.

[0080]

[0081] In one embodiment, the weight vector ω is as follows.

[0082] ω = [0.45 0.24 0.11 0.11 0.045 0.045] T

[0083] Among them, the weight vector indicates that the weight of the number of times N that the specified sub-area is the target area is 0.45, the weight of the sub-area level L(b) is 0.24, the weight of the minimum distance min(d s ) between the sub-area and the base station node is 0.11, the weight of the maximum distance max(d s ) between the sub-area and the base station node is 0.11, the weight of the minimum distance min(d source ) between the sub-area and the source node is 0.045, and the weight of the maximum distance max(d source ) between the sub-area and the source node is 0.045. Then, the consistency test is performed on the weight vector, and the consistency index CR < 1 is obtained, so the weight vector ω meets the consistency requirement.

[0084] S312. Based on the discrimination coefficient, the minimum absolute difference between the evaluation indicators of the sub-area and the evaluation indicators of the reference sequence, and the maximum absolute difference between the evaluation indicators of the sub-area and the evaluation indicators of the reference sequence, the grey correlation coefficient is determined through the grey correlation coefficient calculation formula.

[0085] In this embodiment, the evaluation object is a sub-region. If there are g evaluation objects (i.e., the number of sub-regions is g) and n evaluation indicators, the comparison sequence is x i ={x i (k)}, and the reference sequence is x 0 ={x 0 (k)}; where k = 1, 2,..., n; i = 1, 2,..., g.

[0086] In one embodiment, when the source node selects a target area for any data packet, the sub-regions that have not been designated as target areas have the highest priority, i.e., N = 0; the level L(b) of the target area is located in the middle of the sub-region where the base station node is located and the residential area, i.e., L d / 2, where L d represents the level of the residential area; the closest distance min(d s ) and the farthest distance max(d s ) between the target area and the base station node are both D sd / 2, where D sd is the average distance between the base station node and each residential area; the average of the closest distance min(d source ) and the farthest distance max(d source ) between the target area and the source node is D ss -(D sd / 2), D ss is the Euclidean distance between the base station node and the source node; then the optimal reference sequence is x 0 =(0, L d / 2, L sd / 2, D ss -(L sd / 2), D sd / 2, D ss -(L sd / 2)).

[0087] In this embodiment, the calculation formula of the grey correlation coefficient is:

[0088]

[0089] where ξ i (k) is the grey correlation coefficient between the k-th evaluation indicator of the i-th evaluation object in the comparison sequence and the k-th evaluation indicator of the reference sequence; is the discrimination coefficient, For example is the minimum absolute difference between the k-th evaluation indicator of all evaluation objects and the k-th evaluation indicator of the reference sequence, It is the maximum absolute difference between the k-th evaluation index of all evaluation objects and the k-th evaluation index of the reference sequence.

[0090] S313. Based on the weight vector and the grey correlation coefficient, determine the grey correlation degree value corresponding to each sub-region through the grey correlation degree value calculation formula.

[0091] In this embodiment, the grey correlation degree value calculation formula is:

[0092]

[0093] where r i is the grey correlation degree value of the i-th sub-region, ω(k) is the weight vector of the k-th sub-region, and ξ i (k) is the grey correlation coefficient between the k-th evaluation index of the i-th evaluation object in the comparison sequence and the k-th evaluation index of the reference sequence.

[0094] S314. Sort the grey correlation degree values corresponding to each sub-region, and determine that the sub-region corresponding to the maximum grey correlation degree value is the target region of the data packet.

[0095] In one embodiment, in step S32, based on the neighbor table corresponding to each sensor node and the residence area corresponding to each autonomous underwater vehicle, determine whether any autonomous underwater vehicle obtains a data packet, and through multi-constrained routing node transmission and multi-autonomous underwater vehicle cooperative transmission, transmit the data packet from the source node through each sensor node and each autonomous underwater vehicle to the base station node, including:

[0096] S321. Based on the residence area corresponding to each autonomous underwater vehicle, if the data packet sent by the source node passes through the residence area corresponding to any autonomous underwater vehicle during the transmission process and the autonomous underwater vehicle in the residence area is collecting data packets, the autonomous underwater vehicle in the residence area obtains the data packet, transmits the data packet to the target area, and then transmits it to the base station node through each sensor node.

[0097] S322. Based on the neighbor table corresponding to each sensor node, if the data packet sent by the source node is not obtained by any autonomous underwater vehicle, based on the neighbor table corresponding to each sensor node, the source node sends the data packet to the best next-hop sensor node through the multi-constrained routing node transmission model until the data packet is transmitted to the base station node.

[0098] In one embodiment, after the AUV transmits the data packet to the target area, the AUV will move to a new residence area, so as to achieve dynamic changes and avoid being tracked by attackers. Let the number of the current residence area of the AUV be Num (1 ≤ Num ≤ |B|), then the number of the new residence area is:

[0099] NewIndex(Num) = (Num mod |B|) + 1

[0100] Where Numindex is the number of the new residential area, |B| is the number of residential areas, and Num is the number of the current residential area.

[0101] In one embodiment, the source node is the initial sending node, and the candidate nodes are the neighbor nodes of the sending node. As Figure 4 shown, in step S322, based on the neighbor table corresponding to each sensor node, the source node sends the data packet to the sensor node of the best next hop through the multi-constraint routing node transmission model, including:

[0102] S3221, based on the neighbor table corresponding to each sensor node, determine the set of candidate nodes of the sending node.

[0103] S3222, between each sensor node, determine the cost function value corresponding to each candidate node through the multi-constraint routing node transmission model.

[0104] S3223, sort the cost function values of each candidate node, and determine the candidate node corresponding to the minimum cost function value as the sensor node of the best next hop.

[0105] S3224, the sending node sends the data packet to the sensor node of the best next hop.

[0106] In this embodiment, since the attacker's search range is expanded through irregular multi-path routing, the protection of the source node location privacy is enhanced.

[0107] In one embodiment, as Figure 5 shown, determining the cost function value corresponding to each candidate node in step S3222 through the multi-constraint routing node transmission model includes:

[0108] S32221, based on the forwarding angle constraint value, depth difference constraint value, overhead constraint value, preferred candidate node quantity constraint value, forwarding angle weight coefficient, depth difference weight coefficient, overhead weight coefficient, and preferred candidate node quantity weight coefficient, determine the penalty function value of the candidate node through the penalty function calculation formula;

[0109] S32222, based on the initial energy of the candidate node, the remaining energy of the candidate node, and the selection frequency of the candidate node, determine the objective function value of the candidate node through the objective function calculation formula;

[0110] S32223, based on the objective function value of the candidate node, the penalty function value of the candidate node, the objective function weight coefficient, and the penalty function weight coefficient, determine the cost function value of the candidate node through the cost function calculation formula.

[0111] In this embodiment, due to the constraints on the remaining energy and the number of preferred candidate nodes, it is possible to prevent nodes from dying prematurely due to energy depletion or delivering data packets to areas with sparse node deployments, thus avoiding routing problems in open areas. To avoid long detour problems, constraints are imposed on the depth difference and the forwarding angle to ensure that data packets are always forwarded to nodes closer to the base station node. Finally, constraints on the remaining energy and overhead are imposed, extending the network lifetime. In summary, the MCMR (Multi-Constraint-based Multipath Routing) algorithm effectively protects the location privacy of source nodes, avoids routing problems in open areas, prevents long detour problems, and extends the network lifetime.

[0112] In one embodiment, the network parameters of any sensor node include the forwarding angle, depth difference, overhead, number of preferred candidate nodes, remaining energy, and selection frequency.

[0113] In one embodiment, as Figure 6 shown, Figure 6 is a schematic diagram of the forwarding angle and the forwarding angle constraint value. Set the coordinates of the sending node i as (x i , y i , z i ), and the coordinates of the candidate node j as (x j , y j , z j ); the forwarding angle μ is the angle between the first line segment and the second line segment. The first line segment is the line segment between the base station node and the sending node, and the second line segment is the line segment between the sending node and the candidate node. The candidate nodes include preferred candidate nodes and ordinary candidate nodes; when the forwarding angle μ is smaller, the neighbor nodes of the sending node are closer to the direction where the base station node is located, which can avoid the situation of too long detours. The depth difference Δd is the depth difference between the sending node and the candidate node, Δd = z i - z j , and when the candidate node is between the sending node and the base station node, the greater the depth difference, the closer the candidate node is to the water surface.

[0114] In one embodiment, the overhead is the energy consumption of the sending node to transmit data packets to the candidate node. Specifically, the overhead for transmitting a data packet with a length of l bits is the sum of the energy consumption of the sending node i to send the data packet and the receiving node j to receive the data packet. The calculation formula for the overhead cost(i, j) is:

[0115] cost(i,j) = E t (l,dis(i,j)) + E r (l)

[0116] where, E r(l) represents the energy consumption of the receiving node for receiving data packets, E t (l, dis(i, j)) represents the energy consumption of the sending node for sending data packets; their calculation formulas are respectively:

[0117] E r (l) = lP r T t

[0118] E t (l, dis(i, j)) = lP r T t A(dis(i, j), f)

[0119] Among them, P r represents the power consumption, T t represents the transmission delay; A(dis(i, j), f) represents the underwater acoustic signal attenuation of the acoustic signal with frequency f transmitting the distance dis(i, j). The calculation formula of the underwater acoustic signal attenuation is:

[0120] A(d, f) = d k α(f) d

[0121] Among them, dis(i, j) represents the distance between the sending node and the receiving node. f represents the carrier frequency, with the unit of kHz. k represents the energy diffusion coefficient, and the value of k is different in different scenarios. For example, k = 1.5. α(f) represents the absorption coefficient, with the unit of dB / km. The calculation formula of the absorption coefficient α(f) is:

[0122]

[0123] Among them, f represents the carrier frequency.

[0124] In one embodiment, the number of preferred candidate nodes is preferably the number of neighbor nodes with a depth less than that of the sending node among the neighbor nodes within one hop adjacent to the sending node. The ordinary candidate nodes are the neighbor nodes with a depth greater than or equal to that of the sending node among the neighbor nodes within one hop adjacent to the sending node; preferentially selecting candidate nodes with a large number of preferred candidate nodes to participate in data forwarding can effectively avoid routing problems in open areas. The remaining energy is the remaining energy of the candidate node; preferentially selecting nodes with high remaining energy during the routing process can achieve multi-path routing and balance the energy consumption in the network. Select the frequency as the number of times the candidate node is selected as the best next hop, and preferentially selecting candidate nodes with a low selection frequency can achieve multi-path routing to protect the location privacy of the source node.

[0125] In one embodiment, set the sending node as i, the candidate node as j, the set of candidate nodes as A, and the set of preferred candidate nodes as A 1 , and the set of ordinary candidate nodes as A2 When A 1 is not empty, A = A 1 When A 1 is empty, A = A 2 ; As Figure 6 shown, the constraint condition for the forwarding angle is μ ≦ v; the forwarding angle constraint value v is the included angle between the third line segment between the mapping point of the base station node and the sending node and the first line segment between the sending node and the base station node, so as to prevent the data packet from being transmitted in the direction away from the base station node and avoid the situation of long detours.

[0126] In one embodiment, the calculation of the overhead constraint value is:

[0127]

[0128] where m represents the candidate node.

[0129] In one embodiment, the calculation formula for the preferred candidate node quantity constraint value is:

[0130]

[0131] where m represents the candidate node and n(j) is the preferred candidate node quantity constraint value.

[0132] In one embodiment, the calculation formula for the depth difference constraint value is:

[0133]

[0134] where Δd(i, j) is the depth difference constraint value.

[0135] In one embodiment, the calculation formula for the penalty function value is:

[0136]

[0137] where α is the forwarding angle weight coefficient, β is the overhead weight coefficient, τ is the preferred candidate node quantity weight coefficient, γ is the depth difference weight coefficient, and α + β + τ + γ = 1.

[0138] In one embodiment, the objective function calculation formula is:

[0139]

[0140] where g(x) is the objective function value, E init is the initial energy of the candidate node, E R (j) is the remaining energy of the candidate node, and Num(j) is the selection frequency of the candidate node.

[0141] In one embodiment, the cost function calculation formula is:

[0142] f(x) = ηg(x) + ξp(x)

[0143] Among them, f(x) is the cost function value of the candidate node;

[0144] g(x) is the objective function value of the candidate node; p(x) is the penalty function value of the candidate node;

[0145] η is the objective function weight coefficient; ξ is the penalty function weight coefficient, and η + ξ = 1.

[0146] In one embodiment, set the simulation parameters of the UANs. The side length of the simulation area network is 800m × 800m, the depth of the simulation area is 800m, the number of sensor nodes is 500, the communication radius is 200m, the topology structure is randomly deployed, the data packet size is 1024 bits, the control packet size is 100 bits, the data generation rate is 100 data packets per round, the initial energy of the node is 10J, the data packet transmission power is 10 -5 W, the data sending rate is 10 kbps, the underwater acoustic bandwidth is 20 kHz, the initial energy of the AUV is full load, the speed of the AUV is 8 m / s, the unit power consumption of the AUV is 5 J / m, the initial energy of the attacker is infinite, and the listening range of the attacker is 200m. For the UANs with the above simulation parameters, compare the LDSLP-MA protection method with other protection methods such as SSLP (an underwater acoustic sensor network method for hierarchical source node location privacy protection), PP-SLPP (a probability method for source node location privacy protection in underwater acoustic sensor networks based on push), CS (a low-latency deep-sea AUV data acquisition method), and HAMA (an efficient multi-AUV data acquisition method for underwater sensor networks) from four aspects: safety period, end-to-end delay, average node energy consumption, and average AUV energy consumption. Among them, the end-to-end delay is the average delay of the data packet from the source node to the base station node, and the calculation formula of the end-to-end delay is:

[0147]

[0148] Among them, T sj represents the time when the source node sends the data packet; T rj represents the time when the Sink node successfully receives the data packet; N sink represents the number of data packets successfully received by the Sink node.

[0149] The influence of the objective function weight coefficient η on the protection method. In the MCMR algorithm, the magnitude of the objective function weight coefficient η has a great influence on the performance of the LDSLP-MA method. Figure 7It is a schematic diagram of the influence of the objective function weight coefficient η on the safe period. It can be seen from the figure that as the objective function weight coefficient η increases, the safe period gradually increases and tends to be stable. When the objective function weight coefficient η increases, the sending node is more inclined to select candidate nodes with high remaining energy and low selection frequency as the best next hop, which makes the data transmission path more diversified and makes it more difficult for attackers to find the location of the source node. Figure 8 It is a schematic diagram of the influence of the objective function weight coefficient η on the end-to-end delay. It can be seen from the figure that as the objective function weight coefficient η increases, the end-to-end delay gradually increases. When the objective function weight coefficient η increases, the influence of network parameters such as the forwarding angle and depth difference on the cost function becomes smaller. Therefore, the data packet may reach the base station node through a long detour, resulting in a long end-to-end delay.

[0150] The influence of the communication radius on the protection method. In a three-dimensional UANs, the communication radius of the node has a great influence on data transmission. Figure 9 It is a schematic diagram of the influence of the communication radius on the safe period. It can be seen from the figure that as the communication radius increases, the safe period gradually decreases. When the communication radius becomes larger, the number of hops of the data packet from the source node to the base station node becomes smaller, making it easier for the attacker to track the location of the source node. Even most data packets directly skip the residence area where the AUV is located, resulting in the AUV not participating in the data packet transmission. Therefore, the diversification of the data transmission path is weakened. Therefore, as the communication radius increases, the safe period gradually decreases. Figure 10 It is a schematic diagram of the influence of the communication radius on the end-to-end delay. It can be seen from the figure that as the communication radius increases, the end-to-end delay gradually decreases. When the communication radius increases, the number of hops of the data packet from the source node to the base station node decreases and the participation of the AUV in data transmission is reduced. Therefore, the end-to-end delay also decreases accordingly. Figure 11 It is a schematic diagram of the influence of the communication radius on the average node energy consumption. As the communication radius of the node increases, the opportunity for the AUV to participate in data transmission gradually becomes smaller, and the energy consumption required for data transmission is mostly borne by the nodes in the network. In addition, as the communication radius increases, the hop distance increases, increasing the energy consumption of the nodes. Therefore, it can be seen from the figure that as the communication radius increases, the average node energy consumption gradually increases. Figure 12 It is a schematic diagram of the influence of the communication radius on the average AUV energy consumption. As the communication radius of the node increases, the data packet may directly skip the residence area of the AUV, thereby reducing the participation of the AUV in data transmission. Therefore, as the communication radius increases, the average AUV energy consumption gradually decreases.

[0151] In one embodiment, for the UANs with the above simulation parameters, the LDSLP-MA protection method is compared with other protection methods such as SSLP, PP-SLPP, CS, and HAMA from four aspects: the safe period, the end-to-end delay, the average node energy consumption, and the average AUV energy consumption.

[0152] Among them, from the perspective of the safe period, Figure 13 It is a schematic diagram comparing the number of nodes of five protection methods with the safe period, Figure 14 It is a schematic diagram comparing the depth of five protection methods with the safe period, Figure 15 It is a schematic diagram comparing the network side length of five protection methods with the safe period. The results show that the LDSLP-MA method is significantly better than the other four methods in terms of the safe period. The LDSLP-MA method not only introduces the MCMR algorithm but also proposes a new type of AUV scheduling method to create more diverse and dispersed transmission paths to protect the source node location privacy. It can be seen from these three figures that the safe periods of both the SSLP and PP-SLPP methods are less than that of the LDSLP-MA method. In the SSLP and PP-SLPP methods, the AUV collects and transmits data packets within a specific area, while in the LDSLP-MA method, the AUV travels back and forth in different areas to collect and transmit data, breaking the spatio-temporal correlation of data transmission between regions and expanding the search range of attackers, thus enhancing the security of the source node location privacy. Due to the multi-path technology used in PP-SLPP, its safe period is longer than that of SSLP. The CS and HAMA methods do not use any source node location privacy protection technology. Therefore, compared with the other three methods, the safe periods of these two methods are relatively short. Figure 13 In [reference], since the AUVs in the SSLP and HAMA methods have fixed movement trajectories, the safe periods of these two methods almost remain unchanged as the number of nodes increases. In the CS method, the AUV collects data from the pipeline, and the safe period slowly increases as the number of nodes increases. In PP-SLPP, the number of location information pushed each time the simulation experiment is run is not fixed, resulting in a large change in the distance following the AUV's movement. Therefore, its safe period fluctuates with the change in the number of nodes. As the number of nodes increases, the selection of the next-hop node in the LDSLP-MA method becomes more diverse. Therefore, the safe period of the LDSLP-MA method increases as the number of nodes increases. Figure 14 It shows that the safe periods of the five methods all increase as the depth increases. When the depth increases, the attacker needs to move a longer distance to find the location of the source node, so the safe period also increases accordingly. Figure 15 It shows that as the network side length increases, the safe periods of the five methods gradually increase. Among them, for the three methods of SSLP, CS, and HAMA, as the network side length grows, the safe periods increase slowly, while the safe periods of the LDSLP-MA and PP-SLPP methods have a good growth rate.

[0153] Among them, from the perspective of the end-to-end delay, Figure 16 It is a schematic diagram comparing the number of nodes of five methods with the end-to-end delay, Figure 17 It is a schematic diagram comparing the depth of five methods with the end-to-end delay,Figure 18 It is a schematic diagram comparing the network side length and the end-to-end delay for five methods. It can be seen from these three figures that among these five methods based on multi-AUV cooperation, the LDSLP-MA method achieves the lowest end-to-end delay, and significantly outperforms the end-to-end delays of the PP-SLPP and SSLP methods that protect the source node location privacy based on multi-AUV cooperation. Specifically, since the end-to-end delay of PP-SLPP includes the time for the main AUV to wait for all following AUVs to collect data, and the AUVs in SSLP only collect and transmit data packets along the trajectory, the end-to-end delay of PP-SLPP is greater than that of SSLP. The CS and HAMA methods do not involve waiting for following AUVs to collect data, and the area where AUVs collect data is relatively limited, so their delays are shorter than those of the PP-SLPP and SSLP methods. Although the LDSLP-MA method also relies on AUVs for data transmission, the AUVs do not need to cruise when collecting data packets in this method. In addition, LDSLP-MA allows data packets not collected by AUVs to reach the Sink node through multi-hop transmission, thus avoiding the delay of data packets waiting for AUV collection. Therefore, the end-to-end delay of LDSLP-MA is significantly lower than that of the other four methods based on multi-AUV cooperation. In Figure 16 , as the number of nodes changes, the end-to-end delay of PP-SLPP fluctuates. The fluctuation of the end-to-end delay is caused by the different positions of the AUV cluster push and the initial position of the AUV cluster in PP-SLPP. In SSLP, as the number of nodes increases, the AUV needs to collect data packets from more nodes along the fixed trajectory, resulting in an increase in the end-to-end delay. When the number of nodes increases, HAMA collects data packets from fixed nodes on the fixed trajectory, and there are no deployed nodes in CS, so the end-to-end delays of these two methods remain almost unchanged. In the LDSLP-MA method, the increase in the number of nodes allows for a better choice of the best next-hop node, so the end-to-end delay decreases. Figure 17 It shows that as the depth increases, the end-to-end delay gradually increases. Figure 18 It shows that as the network side length increases, the end-to-end delays of the five methods gradually increase. The increase in the end-to-end delay is due to the fact that the increase in depth or network side length causes the data packets to take longer to reach the base station node from the source node.

[0154] Among them, from the perspective of the average node energy consumption, since there are no sensor nodes deployed in the CS method, Figure 19 It is a schematic diagram comparing the number of nodes and the average node energy consumption of four methods, Figure 20 It is a schematic diagram comparing the depth and the average node energy consumption of four methods, Figure 21It is a schematic diagram comparing the network side length and average node energy consumption of four methods. From these three figures, it is not difficult for us to see that the LDSLP-MA method achieves the lowest average node energy consumption. Specifically, in SSLP, both nodes and AUVs participate in data transmission, and the false packet injection technology is adopted, so its energy consumption is the highest. PP-SLPP also adopts the false packet injection technology, but in PP-SLPP, nodes only push location information to the main AUV and do not participate in data forwarding. Therefore, the average node energy consumption of PP-SLPP is lower than that of SSLP. Since the k-means algorithm used in PP-SLPP for clustering has poor stability, re-clustering requires additional energy consumption. Although ordinary sensor nodes in the PP-SLPP method do not participate in data transmission, false data packets and maintaining the cluster structure still consume a certain amount of energy. In the HAMA method, the AUV moves along a cylindrical curve, and a large number of nodes on this path interact with the AUV to generate data. Therefore, the average node energy consumption of the HAMA method is slightly greater than that of the PP-SLPP method. Since the LDSLP-MA method does not use any energy-wasting measures, and in addition, the AUV undertakes part of the energy consumption, its average node energy consumption is the lowest. In Figure 19 As the number of nodes increases, the number of nodes participating in data transmission in the LDSLP-MA and HAMA methods does not change significantly. Therefore, the average node energy consumption of these two methods does not change significantly. In the PP-SLPP method, the number of location information pushed in each simulation is different, resulting in fluctuations in the average node energy consumption. As the number of nodes increases, SSLP requires more energy consumption for clustering and maintaining the cluster. In addition, nodes are also responsible for transmitting true and false data packets within the cluster. Therefore, its average node energy consumption gradually increases. It can be seen from the figure that as the depth or network side length increases, the average energy consumption of nodes increases. The increase in depth and network side length results in more sparse nodes participating in data transmission or more energy consumption required to maintain the cluster. Therefore, the average node energy consumption increases.

[0155] Among them, from the perspective of average AUV energy consumption, Figure 22 It is a schematic diagram comparing the number of nodes and average AUV energy consumption of five methods, Figure 23 It is a schematic diagram comparing the depth and average AUV energy consumption of five methods, Figure 24Schematic diagram of the comparison between the network side length and the average AUV energy consumption of five methods. The SSLP and HAMA methods collect and transmit data packets along a fixed trajectory. The AUV needs to travel a longer distance to complete the collection and transmission of data packets, resulting in relatively high average AUV energy consumption for these two methods. The randomness of the push position in the PP-SLPP method causes fluctuations in the average AUV energy consumption. In the LDSLP-MA method, each AUV continuously wanders in different regions to transmit data packets, resulting in a longer travel distance of the AUV than that of PP-SLPP. Therefore, the average AUV energy consumption of LDSLP-MA is higher than that of PP-SLPP. In the CS method, the AUV collects data packets in the pipeline along a predetermined circular route, so its average AUV energy consumption is the lowest. Although the average AUV energy consumption of the LDSLP-MA method is slightly higher than that of methods such as CS and PP-SLPP, due to the mobility and flexibility of the AUV, it is easier to charge compared to sensor nodes. In Figure 22 In the LDSLP-MA, PP-SLPP, CS, and HAMA methods, the number of nodes does not affect the travel distance of the AUV. Therefore, as the number of nodes increases, the average AUV energy consumption of these methods remains almost unchanged. In the SSLP method, as the number of nodes increases, the AUV needs to travel to collect data from more nodes, resulting in an increase in the average AUV energy consumption. As the depth or network side length increases, the AUV needs to sail a longer distance, so the average AUV energy consumption gradually increases.

[0156] After a series of simulation experiments, it is verified that the LDSLP-MA method has lower end-to-end delay, higher source node location privacy security, and smaller energy consumption compared with other multi-AUV cooperation-based methods. Different from methods such as SSLP and PP-SLPP that sacrifice delay to achieve source node location privacy protection, the LDSLP-MA method achieves source node location privacy protection without sacrificing delay. Specifically, the LDSLP-MA method has good performance in terms of performance indicators such as safety period, end-to-end delay, average node energy consumption, and average AUV energy consumption. The simulation experiments show that the LDSLP-MA method is a low-delay multi-AUV cooperation-based source node location privacy protection method. The performance of the LDSLP-MA method in various performance indicators is as follows: the safety period is excellent, the end-to-end delay is excellent, the average node energy consumption is excellent, and the average AUV energy consumption is medium.

[0157] This application provides a novel method for protecting the location privacy of source nodes in multi-AUV cooperative underwater acoustic networks (UANs), called the LDSLP-MA method, to address the common long-delay problem in existing methods for protecting the location privacy of source nodes in multi-AUV cooperative UANs. This method has two key innovations: the MCMR algorithm and a low-delay multi-AUV scheduling method. In the MCMR algorithm, multi-path routing is created by considering the remaining energy and selection frequency of candidate nodes, thereby enhancing the security of source node location privacy. In addition, the network performance is optimized by adding constraints to network parameters such as the forwarding angle, depth difference, overhead, and the number of preferred candidate nodes of candidate nodes. To alleviate the high delay caused by AUV cruising, the residence area and target area of AUVs are reasonably planned. This scheduling method can not only reduce the delay but also diversify the data packet transmission path, thereby enhancing the security of source node location privacy. Simulation results show that, compared with other multi-AUV cooperative methods, the LDSLP-MA method exhibits superior performance in terms of the security period, end-to-end delay, and energy consumption. These simulation results demonstrate the effectiveness of the LDSLP-MA method in protecting the location privacy of source nodes and reducing end-to-end delay in UANs.

[0158] It should be understood that the sequence numbers of the steps in the above embodiments do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0159] The above embodiments are only used to illustrate the technical solutions of this application, rather than to limit it; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included in the protection scope of this application.

Claims

1. A method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multi-AUV collaboration, characterized in that: include: The three-dimensional underwater acoustic network is divided into multiple three-dimensional sub-areas by using the k-dimensional tree method. The underwater acoustic network includes multiple underwater sensor nodes, multiple autonomous underwater vehicles and a base station node on the water surface. Any underwater sensor node is a source node. Based on each sub-area and each sensor node, determining the hierarchical information corresponding to each sub-area, the neighbor table corresponding to each sensor node, and the residential area corresponding to each host underwater vehicle; Based on the hierarchical information corresponding to each sub-area, the neighbor table corresponding to each sensor node and the residential area corresponding to each autonomous underwater vehicle, the data packet is transmitted from the source node through each sensor node and each autonomous underwater vehicle to the base station node through multi-constrained routing node transmission and multi-autonomous underwater vehicle collaborative transmission. The residential area is the area where the autonomous underwater vehicle collects the data packet.

2. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multiple AUVs as described in claim 1 is characterized in that: The k-dimensional tree method is used to divide the three-dimensional hydroacoustic network into multiple three-dimensional sub-areas, including: Establish a three-dimensional coordinate system with the base station node as the origin, and obtain the three-dimensional coordinates of multiple data points; Determine the variance value of one dimension based on the coordinates of any data point in one dimension, the average coordinates of each data point in one dimension, and the number of data points; The dimension corresponding to the maximum variance value is used as the split axis, and the median coordinates of each data point on the split axis are determined as the split point; Based on the dimension of the split axis, data points with values ​​less than the median coordinate are divided into the left branch, and data points with values ​​greater than the median coordinate are divided into the right branch; The left branch and the right branch are recursively constructed using the three dimensions as the partitioning axes in turn until the hydroacoustic network is divided into multiple three-dimensional sub-regions.

3. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multi-AUV collaboration as claimed in claim 1, characterized in that: The target area is the destination of the data packet transmitted by the autonomous underwater vehicle; Based on the hierarchical information corresponding to each sub-area, the neighbor table corresponding to each sensor node and the residential area corresponding to each master underwater vehicle, the data packet is transmitted from the source node through each sensor node and each master underwater vehicle to the base station node through multi-constrained routing node transmission and multi-autonomous underwater vehicle cooperative transmission, including: Determine the target area of ​​the data packet based on the level information corresponding to each sub-area and the gray correlation value corresponding to each sub-area; Based on the neighbor table corresponding to each sensor node and the residential area corresponding to each autonomous underwater vehicle, it is determined whether any autonomous underwater vehicle obtains the data packet. Through multi-constrained routing node transmission and multi-autonomous underwater vehicle collaborative transmission, the data packet is transmitted from the source node through each sensor node and each autonomous underwater vehicle to the base station node.

4. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multi-AUV collaboration as claimed in claim 3 is characterized in that: Based on the level information corresponding to each sub-region and the gray correlation value corresponding to each sub-region, the target area of ​​the data packet is determined, including: Determine the weight vector corresponding to each evaluation index of the sub-region by means of hierarchical analysis, wherein the evaluation indexes include the sub-region level, the number of times the sub-region is designated as the target region, the closest distance between the sub-region and the base station node, the farthest distance between the sub-region and the base station node, the closest distance between the sub-region and the source node, and the farthest distance between the sub-region and the source node; Based on the discrimination coefficient, the minimum absolute difference between the evaluation index of the sub-region and the evaluation index of the reference sequence, and the maximum absolute difference between the evaluation index of the sub-region and the evaluation index of the reference sequence, the grey correlation coefficient is determined by the grey correlation coefficient calculation formula; Based on the weight vector and the grey correlation coefficient, the grey correlation value corresponding to each sub-region is determined by the grey correlation value calculation formula; The grey correlation values ​​corresponding to the sub-regions are sorted, and the sub-region corresponding to the maximum grey correlation value is determined as the target region of the data packet.

5. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network of multiple AUVs as claimed in claim 3, characterized in that: Based on the neighbor table corresponding to each sensor node and the residential area corresponding to each autonomous underwater vehicle, it is determined whether any autonomous underwater vehicle obtains a data packet, and the data packet is transmitted from the source node through each sensor node and each autonomous underwater vehicle to the base station node through multi-constrained routing node transmission and multi-autonomous underwater vehicle cooperative transmission, including: Based on the residential areas corresponding to each autonomous underwater vehicle, if the data packet sent by the source node passes through the residential area corresponding to any autonomous underwater vehicle and the autonomous underwater vehicle in the residential area is collecting data packets, the autonomous underwater vehicle in the residential area obtains the data packet and transmits the data packet to the target area, and then transmits it to the base station node through each sensor node; Based on the neighbor table corresponding to each sensor node, if the data packet sent by the source node is not obtained by any autonomous underwater vehicle, the source node sends the data packet to the best next-hop sensor node through the multi-constraint routing node transmission model until the data packet is transmitted to the base station node.

6. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multiple AUVs as claimed in claim 5, characterized in that: Based on the neighbor table corresponding to each sensor node, the source node sends the data packet to the best next-hop sensor node through a multi-constrained routing node transmission model, including: Based on the neighbor table corresponding to each sensor node, the candidate node set of the sending node is determined, the source node is the initial sending node, and the candidate node is the neighbor node of the sending node; Between each sensor node, the cost function value corresponding to each candidate node is determined through a multi-constrained routing node transmission model; Sort the cost function values ​​of each candidate node, and determine the candidate node corresponding to the minimum cost function value as the best next-hop sensor node; The sending node sends the data packet to the sensor node with the best next hop.

7. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multiple AUVs as claimed in claim 6 is characterized in that: The cost function value corresponding to each candidate node is determined through a multi-constrained routing node transmission model, including: Based on the forwarding angle constraint value, the depth difference constraint value, the overhead constraint value, the preferred candidate node quantity constraint value, the forwarding angle weight coefficient, the depth difference weight coefficient, the overhead weight coefficient and the preferred candidate node quantity weight coefficient, the penalty function value of the candidate node is determined by the penalty function calculation formula; Based on the initial energy of the candidate node, the residual energy of the candidate node and the selection frequency of the candidate node, the objective function value of the candidate node is determined by the objective function calculation formula; Based on the objective function value of the candidate node, the penalty function value of the candidate node, the objective function weight coefficient and the penalty function weight coefficient, the cost function value of the candidate node is determined by a cost function calculation formula.

8. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multiple AUVs as claimed in claim 7, characterized in that: The cost function calculation formula is: f(x)=ηg(x)+ξp(x) Among them, f(x) is the cost function value of the candidate node; g(x) is the objective function value of the candidate node; p(x) is the penalty function value of the candidate node; η is the weight coefficient of the objective function; ξ is the weight coefficient of the penalty function, η+ξ=1.

9. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multiple AUVs as claimed in claim 7, characterized in that: The network parameters of any sensor node include forwarding angle, depth difference, overhead, number of preferred candidate nodes, remaining energy, and selection frequency; The forwarding angle is the angle between the first line segment and the second line segment, the first line segment is the line segment between the base station node and the sending node, the second line segment is the line segment between the sending node and the candidate node, and the candidate node includes a preferred candidate node and a common candidate node; The depth difference is the depth difference between the sending node and the candidate node; The overhead is the energy consumption of the sending node in transmitting the data packet to the candidate node; The number of preferred candidate nodes is the number of neighbor nodes of the sending node whose depth is less than the number of neighbor nodes of the sending node among the neighbor nodes of the sending node within one hop, and the number of common candidate nodes is the number of neighbor nodes of the sending node whose depth is greater than or equal to the number of neighbor nodes of the sending node among the neighbor nodes of the sending node within one hop; The residual energy is the residual energy of the candidate node; The selection frequency is the number of times a candidate node is selected as the best next hop.

10. The method for protecting the location privacy of source nodes in a low-latency underwater acoustic network with multiple AUVs as claimed in claim 7, characterized in that: The objective function calculation formula is: Among them, g(x) is the objective function value, E init is the initial energy of the candidate node, E R (j) is the residual energy of the candidate node, and Num(j) is the selection frequency of the candidate node.

Images