Positioning method for water surface sensor network under non-line-of-sight condition with unknown signal propagation loss

Abstract

The application relates to a water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss, comprising the following steps: constructing a radio signal strength ranging model under non-line-of-sight conditions with distance constraints, and constructing a joint estimation optimization function according to the ranging model; improving a binary chopping method based on the optimization function, and jointly estimating a path loss factor and a target position; remodeling the optimization function according to the estimated target position, re-optimizing the target position, obtaining an optimized target position, and completing water surface sensor network positioning under non-line-of-sight conditions. The application can adapt to a high-dynamic water surface environment, and solve the problem of increased positioning error caused by unknown path loss factors and non-line-of-sight influence in the process of water surface radio signal propagation in the ocean.

Description

Technical Field

[0001] This invention relates to the field of target node localization technology for marine wireless sensor networks, and in particular to a method for localizing surface sensor networks under non-line-of-sight conditions where signal propagation loss is unknown. Background Technology

[0002] Surface sensor networks (SSNs) are a crucial component of marine three-dimensional monitoring networks. They provide diverse environmental situational awareness information for the safe navigation of intelligent vessels and are an effective technical means to prevent accidents at sea. In SSNs, sensing data is typically considered valid only when it carries location information. Therefore, obtaining relatively accurate target locations, commonly known as positioning, is considered a fundamental requirement for marine monitoring applications.

[0003] However, in marine environments, especially on the water, passing ships often obstruct signals, causing them to propagate beyond line-of-sight. Furthermore, the path loss of signals in the air varies with sea surface temperature and humidity, further complicating SSN positioning and increasing accuracy. Existing positioning technologies lack effective solutions to these problems and cannot achieve high positioning accuracy under non-line-of-sight conditions with unknown signal propagation losses. Summary of the Invention

[0004] The purpose of this invention is to provide a water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss. This method can adapt to highly dynamic water surface environments and achieve high positioning accuracy under non-line-of-sight conditions and with unknown signal propagation loss factors.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] Localization methods for water surface sensor networks with unknown signal propagation loss under non-line-of-sight conditions include:

[0007] A distance-constrained radio received signal strength ranging model under non-line-of-sight conditions is constructed, and a joint estimation optimization function is constructed based on the ranging model;

[0008] The binary chopper method is improved based on the optimization function, and the path loss factor and the target position are jointly estimated.

[0009] Based on the estimated target position, the optimization function is reshaped, and the target position is further optimized to obtain the optimized target position, thus completing the water surface sensor network positioning under non-line-of-sight conditions.

[0010] Furthermore, constructing the distance-constrained radio received signal strength ranging model under non-line-of-sight conditions includes:

[0011] Anchor nodes are set up on the water surface of the sea area to be located, and motion constraint information of the anchor nodes is collected. Based on the motion constraint information, combined with the path loss model of signal propagation and the non-line-of-sight influencing factors of signal propagation, a radio received signal strength ranging model under the distance constraint under the non-line-of-sight condition is constructed.

[0012] Furthermore, the ranging model is constructed as follows:

[0013]

[0014]

[0015] in, This represents the power received by the target node from the i-th anchor node at time t; PL(d0) represents the transmit power of the target node at time t; PL(d0) represents the loss value at a reference distance of d0, where d0 is the reference distance, typically 1m; α t Indicates the path loss factor; Let be the distance between the i-th anchor node and the target node; This represents the corresponding signal attenuation noise; This indicates non-line-of-sight noise during signal propagation caused by the movement of vessels on the water.

[0016] Furthermore, the characteristic is that the motion constraint information of the anchor node includes:

[0017] N anchor nodes and 1 target node are deployed on the surface of the monitored sea area, and the movement of the anchor nodes on the water surface is constrained by... The area within a circle of radius L is defined as follows: L is the length of the anchor chain, and Dep is the water depth; the position of the i-th anchor node at time t is... The position of the target node at time t is: Where T represents transpose.

[0018] Furthermore, the feature is that constructing the joint estimation optimization function includes:

[0019] The ranging model is linearized to construct a joint estimation optimization function with path loss factor and target position as variables:

[0020]

[0021]

[0022] Where, k iThe function expression relating the received signal strength to the introduced parameter v is expressed as follows: wi represents the corresponding weight information, expressed as δ is the functional expression between the introduced parameter and the path loss factor, which can be expressed as δ=(α t -v) / v;||·|| is the second-order norm.

[0023] Furthermore, the improved binary chopper method is characterized by:

[0024] Based on the constraints of the optimization function, the trust region interval is constructed as follows:

[0025] ξ=max{-diag(Γ T ωΓ) / diag(D), λ},

[0026] Where ξ represents the optimal number within the trust region. diag(·) denotes a diagonal matrix, λ is the multiplier, and Γ is the matrix composed of anchor nodes and κ. i The matrix formed is represented as Γ T Let Γ be the transpose of matrix Γ; D is a matrix consisting of 0s, 1s, and the identity matrix I, denoted as .

[0027] Obtain the optimal number within the trust region interval, solve for the multiplier and substitute it back into the trust region interval, thus transforming the original binary chopper interval from {1 / -diag(Γ)} T ωΓ),∞} shrinks to {-diag(Γ) T ωΓ) / diag(D),λ}.

[0028] Further, the feature is that jointly estimating the path loss factor and the target location includes:

[0029] Estimate the target location based on the optimal number within the trust region interval:

[0030]

[0031] in, Let Λ represent the target position vector, and κ be the vector formed by the anchor node and κ. i The vector formed is represented as f is the corresponding vector, denoted as

[0032] Based on the obtained vector elements Further estimate the path loss factor:

[0033]

[0034] Furthermore, the characteristic is that reshaping the optimization function includes:

[0035] Based on the estimated target position, an error correction mechanism is introduced to reshape the optimization function as follows:

[0036] Φζ=Ξ+Ω,

[0037] Where Φ is a matrix consisting of 0s and 1s, denoted as: ζ is a vector consisting of the squares of the actual positions, denoted as... Ξ is a vector composed of the squares of the estimated positions, denoted as... They represent the solution vectors respectively. Corresponding serial number element;

[0038]

[0039] Furthermore, the feature is that the target position is further optimized to obtain the optimized target position, including:

[0040] The reshaped optimization function is solved using the weighted least squares method to correct errors and obtain the optimized target position.

[0041]

[0042] Where sgn(·) represents the sign function; for The corresponding first and second elements.

[0043] The beneficial effects of this invention are as follows:

[0044] Compared with existing technologies, the positioning method proposed in this invention can adapt to highly dynamic water surface environments and maintain good positioning performance under non-line-of-sight conditions and unknown signal propagation loss. It solves the problem of increased positioning error caused by unknown path loss factor and non-line-of-sight influence during the propagation of radio signals on the ocean surface. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart of the method in an embodiment of the present invention;

[0047] Figure 2This is a diagram illustrating the restricted motion characteristics of the buoy anchor node in an embodiment of the present invention.

[0048] Figure 3 This refers to the estimation error corresponding to different numbers of anchor nodes in the embodiments of the present invention;

[0049] Figure 4 The estimation errors corresponding to different ranging noises in the embodiments of the present invention;

[0050] Figure 5 This represents the estimation error corresponding to different non-line-of-sight noises in the embodiments of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] This invention provides a method for locating a surface sensor network under non-line-of-sight conditions where signal propagation loss is unknown. The process is as follows: Figure 1 As shown, it includes:

[0054] S1. Based on signal propagation loss and the motion characteristics of surface buoys, construct a distance-constrained radio received signal strength ranging model under non-line-of-sight conditions;

[0055] S2. Construct a joint estimation optimization function with path loss factor and target position as variables through linearization operation;

[0056] S3. Utilize the improved binary chopper method to jointly estimate the path loss factor and target location;

[0057] S4. Based on the estimated target position, reshape the optimization function using an error correction mechanism, and then perform a re-optimization operation on the target position.

[0058] Further optimization of the plan, step S1 specifically includes:

[0059] S11. Deploy N buoy sensor nodes containing GPS location information, namely anchor nodes and 1 target node, on the surface of the monitored sea area.

[0060] Assume the motion of the buoy anchor node on the water surface is constrained by a certain range. Inside a circle with radius , such as Figure 2 As shown, where L is the anchor chain length and Dep is the water depth; assuming the position of the i-th anchor node at time t is... The position of the target node at time t is Where T represents transpose.

[0061] S12. The anchor node and the target node exchange information via radio signals regarding the Signal Strength Indication Value (RSSI). Due to the presence of passing vessels on the water, signal propagation may encounter obstruction, preventing information exchange at line-of-sight. Therefore, considering the path loss model of signal propagation, the constrained motion of the anchor node, and the non-line-of-sight effects of passing vessels on signal propagation, a ranging model is established:

[0062]

[0063] in, This represents the power received by the target node from the i-th anchor node at time t; PL(d0) represents the transmit power of the target node at time t; PL(d0) represents the loss value at a reference distance of d0, where d0 is the reference distance, typically 1m. α t Indicates the path loss factor; Let be the distance between the i-th anchor node and the target node; Let represent the corresponding signal attenuation noise. Assume the noise variance is equal at every moment and follows a mean of zero with a variance of . The Gaussian distribution of can be expressed as This represents the non-line-of-sight noise during signal propagation caused by the movement of vessels on the water. It is assumed that the noise variance is equal at every moment and follows a mean of zero, with a variance of... The Gaussian distribution of can be expressed as

[0064] Step S2 specifically includes:

[0065] S21. By rearranging formula (1) and squaring the distances between each anchor node and the target node at time t, we can obtain:

[0066]

[0067] S22. Using linearization operations, respectively... and Performing a Taylor series expansion, when d0 = 1m, formula (2) can be expressed as:

[0068]

[0069] S23. Introduce the parameter υ to construct an identity, such that αt =υ(1+(α) t -υ) / υ), and let δ=(α) t -υ) / υ, then formula (3) can be transformed into:

[0070]

[0071] S24. A linear expansion of formula (4) yields:

[0072]

[0073] Where, k i The function expression relating the received signal strength to the introduced parameter v is expressed as follows:

[0074] S25. Construct the joint estimation optimization function:

[0075]

[0076] Where, ω i The corresponding weight information is represented as ||·|| is the second-order norm.

[0077] S26, Order Expanding formula (6) and rearranging it, we get:

[0078]

[0079] in, diag(·) represents a diagonal matrix;

[0080]

[0081] Where I and 0 represent the identity matrix and the zero matrix, respectively.

[0082] Step S3 specifically includes:

[0083] S31. Based on the constraints of formula (7), the idea of ​​linear interpolation is introduced to further construct the trust region interval, and the optimization function of formula (7) is solved as follows:

[0084] ξ=max{-diag(Γ T ωΓ) / diag(D),λ}, (9)

[0085] Where ξ represents the optimal number within the trust region; λ is the multiplier.

[0086] S32. Combining the trust region interval and constraints of formula (9) and formula (7), the binary chopper method is further improved to jointly estimate the path loss factor and target position. The specific process includes:

[0087] S32.1. Based on the constraints of formula (7), solve for the multiplier λ using equation (10):

[0088]

[0089] in, Represents the equation corresponding to the multiplier; ο=(Γ T ωΓ+λD) -1 (Γ T ωΛ+λf).

[0090] S32.2 Substitute the multiplier obtained by formula (10) into the trust region interval. The original binary chopper interval is changed from {1 / -diag(Γ)}. T ωΓ),∞} shrinks to {-diag(Γ) T ωΓ) / diag(D),λ}, and the optimal number within the trust region interval is obtained through formula (9).

[0091] S32.3. Substitute the obtained optimal number ξ within the trust region into formula (11) to obtain the variable to be solved:

[0092]

[0093] S32.4, Based on the obtained vector elements Substituting further into formula (12), the path loss factor is estimated:

[0094]

[0095] Step S4 specifically includes:

[0096] S41. After obtaining the estimated position of the target according to formula (11), an error correction mechanism is introduced to reshape the optimization function as follows:

[0097] Φζ=Ξ+Ω, (13)

[0098] Where Φ is a matrix composed of 0s and 1s, which can be represented as ζ is a vector consisting of the squares of the actual positions, which can be represented as: Ξ is a vector composed of the squares of the estimated positions, which can be represented as: They represent the solution vectors respectively. Corresponding serial number element;

[0099]

[0100] S42. Solve formula (13) using the weighted least squares method to correct errors and further optimize target position information:

[0101]

[0102] S43. The value obtained by solving formula (15) is the corresponding variable after squaring. Formula (16) is used to obtain the optimized target position sign and precise value:

[0103]

[0104] Where sgn(·) represents the sign function; for The corresponding first and second elements.

[0105] In this embodiment, to verify the effectiveness of the proposed algorithm (NUPLL), simulation experiments were conducted in Matlab R2018b. For different scenarios, the algorithm was compared with existing algorithms LS, TRL, and MFMM. The minimum root mean square error (RSME) was used as the evaluation metric to assess the algorithm's performance.

[0106]

[0107] Where MC represents the total number of Monte Carlo simulations; nu represents the number of Monte Carlo simulations; x t Indicates the actual location of the target; This indicates the estimated location of the target.

[0108] Furthermore, to simulate the dynamic scene of the water surface in the ocean, the positions of the anchor node and the target node are set to change in each Monte Carlo simulation, with the anchor chain length of the anchor node set to 25m and the water depth set to 15m. Therefore, the water surface area where the anchor node's movement is restricted is within a circle with a radius of 20m. Other fixed parameters are set as follows: υ = 5.

[0109] Figure 3 Indicating ranging noise Non-line-of-sight noise In this case, the estimation error of the target location varies with the number of anchor nodes. According to... Figure 3 The information shows that the positioning accuracy improves with the increase in the number of anchor nodes, because the increased number of anchor nodes leads to more usable positioning information. Furthermore, good positioning accuracy is maintained even when the number of buoy anchor nodes is relatively small.

[0110] Figure 4 This indicates that with N=10 anchor nodes, non-line-of-sight noise... Under different ranging noise conditions, the estimation error of the target position varies. According to... Figure 4As can be seen from the information, positioning accuracy decreases with increasing ranging noise. The LS algorithm is particularly sensitive to noise, exhibiting the largest increase in positioning error as noise levels rise. While the error increases of MFMM and TRL are smaller than those of LS, their overall positioning accuracy is lower than that of the NUPLL method proposed in this invention. Therefore, when environmental ranging noise increases, the method proposed in this invention is superior to existing methods.

[0111] Figure 5 This indicates that with N=10 anchor nodes, the ranging noise... Under different non-line-of-sight noise conditions, the estimation error of the target position varies. According to... Figure 5 As can be seen from the information, the various methods are relatively robust to non-line-of-sight noise. The NUPLL method proposed in this invention has a smaller positioning error compared to other methods, and can improve positioning accuracy to a certain extent when non-line-of-sight noise increases.

[0112] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for locating a surface sensor network under non-line-of-sight conditions where signal propagation loss is unknown, characterized in that... include: A distance-constrained radio received signal strength ranging model under non-line-of-sight conditions is constructed, and a joint estimation optimization function is constructed based on the ranging model; The improved binary chopper method, based on the aforementioned optimization function and jointly estimated by the path loss factor and target location, includes: Based on the constraints of the optimization function, the trust region interval is constructed as follows: in, This represents the optimal number within the trust region. , This represents a diagonal matrix, where ωi represents the corresponding weight information. For the vehicle, For anchor nodes and The matrix formed is represented as a i t Let be the position of the i-th anchor node at time t, and N be the number of anchor nodes. This represents a functional expression relating the received signal strength to the introduced parameter v. It consists of 0, 1 and the identity matrix The constructed matrix is ​​represented as ; Obtain the optimal number within the trust region interval, solve for the multiplier and substitute it back into the trust region interval, thus transforming the original binary chopper method interval from... Shrink to ; Based on the estimated target position, the optimization function is reshaped, and the target position is further optimized to obtain the optimized target position, thus completing the water surface sensor network positioning under non-line-of-sight conditions.

2. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 1, characterized in that, The construction of the distance-constrained radio received signal strength ranging model under non-line-of-sight conditions includes: Anchor nodes are set up on the water surface of the sea area to be located, and motion constraint information of the anchor nodes is collected. Based on the motion constraint information, combined with the path loss model of signal propagation and the non-line-of-sight influencing factors of signal propagation, a radio received signal strength ranging model under the distance constraint under the non-line-of-sight condition is constructed.

3. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 2, characterized in that, The ranging model is constructed as follows: in, Indicates the first Anchor nodes at The power received from the target node at any given time; Indicates that the target node is in Transmission power at any given moment; Indicates the reference distance as The loss value at that time, This represents a reference distance, typically 1 meter. Indicates the path loss factor; For the first The distance between each anchor node and the target node; This represents the corresponding signal attenuation noise; This indicates non-line-of-sight noise during signal propagation caused by the movement of vessels on the water. and Let x and y represent the x and y coordinates of the i-th anchor node at time t, respectively, and M be the radius of the circle restricting the movement of the anchor node.

4. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 2, characterized in that, The motion constraint information of the anchor node includes: N anchor nodes and 1 target node are deployed on the surface of the monitored sea area, and the movement of the anchor nodes on the water surface is constrained by... Inside a circle with radius , where, The length of the anchor chain. For water depth; Time of the first The positions of the anchor nodes are The target node The position of the moment is ,in, This indicates transpose.

5. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 3, characterized in that, Constructing the joint estimation optimization function includes: The ranging model is linearized to construct a joint estimation optimization function with path loss factor and target position as variables: in, Regarding the received signal strength and introduced parameters v The function expression between them is represented as , This represents the power received by the target node from the i-th anchor node at time t. This represents the transmit power of the target node at time 0; The corresponding weight information is represented as , To introduce a functional expression between the parameters and the path loss factor, it can be expressed as follows: ; It is a second-order norm. For the target node The position at time N represents the number of anchor nodes, and i represents the i-th anchor node. Let t be the position of the i-th anchor node at time t.

6. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 1, characterized in that, The joint estimation of the path loss factor and target location includes: Estimate the target location based on the optimal number within the trust region interval: in, Represents the target position vector. For anchor nodes and The vector formed is represented as , For the corresponding vector, it is represented as , For anchor nodes and The matrix formed This represents a functional expression relating the received signal strength to the introduced parameter v. It consists of 0, 1 and the identity matrix The matrix formed Let N represent the optimal number within the trust region interval, where N is the number of anchor nodes and i is the i-th anchor node. Let be the position of the i-th anchor node at time t; Based on the obtained vector elements Further estimate the path loss factor: 。 7. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 1, characterized in that, The optimization function is reshaped to include: Based on the estimated target position, an error correction mechanism is introduced to reshape the optimization function as follows: in, Let the matrix consist of 0s and 1s, represented as ; Let the vector be composed of the squares of the actual positions, denoted as: ; Let be a vector composed of the squares of the estimated positions, denoted as , , , They represent the solution vectors respectively. Corresponding serial number element; 。 8. The water surface sensor network positioning method under non-line-of-sight conditions with unknown signal propagation loss according to claim 7, characterized in that, Further optimization of the target location to obtain the optimized target location includes: The reshaped optimization function is solved using the weighted least squares method to correct errors and obtain the optimized target position. in, Represents a symbolic function; , for The corresponding first and second elements, , These represent the index elements corresponding to the solution vectors.

Images