A positioning method for ocean sensor network with parameter uncertainty

Abstract

The application relates to a positioning method for an ocean sensor network with uncertain parameters, and the method comprises the following steps: S1, constructing a ranging model based on received signal strength according to signal propagation loss; S2, selecting a reference node, performing a difference operation on the ranging model of the reference node, introducing a natural constant, constructing a differential ranging model, and solving an initial position of a target by using a linear unbiased estimation method; S3, establishing a target optimization function with a path loss factor as a variable to be solved based on the initial position of the target, constructing a constraint, and solving an optimal path loss factor by using an improved whale optimization algorithm; and S4, reconstructing the differential ranging model based on the optimal path loss factor, solving the reconstructed differential ranging model by using a linear least square method, and obtaining an optimized target position. Compared with the prior art, the application has the advantages of high positioning accuracy.

Description

Technical Field

[0001] This invention relates to the field of marine wireless sensor network node positioning technology, and in particular to a marine sensor network positioning method with uncertain parameters. Background Technology

[0002] As a crucial pathway for maritime information interconnection, research on Ocean Sensor Networks (OSNs) holds significant academic and scientific value. It has broad application prospects not only in maritime transportation but also in marine ecological monitoring and maritime search and rescue. One of the key technologies is node positioning, which allows for the acquisition of data location information, providing a reliable basis for subsequent decision-making.

[0003] However, obtaining reliable location information for target nodes in the complex and ever-changing marine environment is a challenge. This is especially true because equipment wear and tear, as well as variations in environmental parameters such as temperature and humidity, lead to significant uncertainties in signal propagation parameters, resulting in decreased positioning accuracy. Existing positioning technologies are all based on known propagation parameters and cannot effectively handle situations with significant uncertainties in these parameters, thus increasing positioning errors. Summary of the Invention

[0004] The purpose of this invention is to overcome the defects of the prior art by providing a marine sensor network positioning method with uncertain parameters, so as to solve the problem of increased positioning error caused by the uncertainty of signal propagation parameters due to equipment wear and tear and changes in environmental parameters.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for locating marine sensor networks with uncertain parameters includes the following steps:

[0007] S1. Construct a ranging model based on the received signal strength according to the signal propagation loss;

[0008] S2. Select a reference node, perform a subtraction operation on the ranging model of the reference node, introduce the natural constant, construct a differential ranging model, and use the linear unbiased estimation method to solve for the initial position of the target.

[0009] S3. Based on the initial target position, establish an objective optimization function with the path loss factor as the variable to be solved, construct constraints, and use the improved whale optimization algorithm to solve for the optimal path loss factor.

[0010] S4. Based on the optimal path loss factor, the differential ranging model is reconstructed, and the reconstructed differential ranging model is solved by the linear least squares method to obtain the optimized target position.

[0011] Furthermore, the ranging model is as follows:

[0012]

[0013] Among them, P ri di represents the received signal strength of the i-th anchor node; d0 represents the reference distance value; P0 represents the target's transmit power; α represents the path loss factor; γ i This represents Gaussian distributed wave-masking noise.

[0014] Furthermore, the path loss factor is set to [2, 6].

[0015] Furthermore, the improved whale optimization algorithm solves the objective optimization function through three steps: surrounding the prey, hunting behavior, and searching for the prey.

[0016] Furthermore, the process of surrounding the prey specifically involves introducing perturbation and random numbers to update the target optimization function.

[0017] Furthermore, the hunting behavior specifically simulates the spiral hunting behavior of whales, and the target optimization function is updated based on the spiral attack hunting trend of whales.

[0018] Furthermore, the whale's spiral attack hunting tendency includes both contraction and spiral attack, each with a probability of 0.5.

[0019] Furthermore, the difference operation involves subtracting the distance measurement model corresponding to other anchor nodes outside the reference node from the distance measurement model of the reference node.

[0020] Furthermore, after performing a subtraction operation on the ranging model of the reference node, the logarithm in the obtained ranging model is subjected to a base change and squaring operation, and a natural constant is introduced after the squaring operation.

[0021] Furthermore, by introducing the natural constant, a differential ranging model is constructed based on the first-order Taylor series expansion.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] (1) Select a reference node and perform a subtraction operation on the ranging model of the reference node. This can eliminate the influence of unknown transmission power, overcome the positioning problem when the transmission power, a propagation parameter, has a large uncertainty, and improve the positioning accuracy.

[0024] (2) Based on the initial position of the target, establish the target optimization function with the path loss factor as the variable to be solved, construct constraints, and use the improved whale optimization algorithm to solve the optimal path loss factor, so as to overcome the positioning problem when the path loss factor, a propagation parameter, has a large uncertainty and improve the positioning accuracy. Attached Figure Description

[0025] Figure 1 This is a flowchart of the present invention;

[0026] Figure 2 Flowchart for solving the optimal path loss factor using the improved whale optimization algorithm;

[0027] Figure 3 The positioning error diagram shows the positioning error under different noise conditions.

[0028] Figure 4 This is a diagram showing the positioning error under different numbers of anchor nodes. Detailed Implementation

[0029] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0030] This invention provides a method for locating marine sensor networks with uncertain parameters. The flowchart of the method is as follows: Figure 1 As shown, the method includes the following steps:

[0031] S1. Based on the signal propagation loss, construct a ranging model based on the received signal strength.

[0032] The specific steps of S1 are as follows:

[0033] Consider an ocean sensor network containing N anchor nodes. If the deployment depth is known, the 3D scene can be transformed into a 2D scene. Assume the position of the i-th anchor node can be represented as a. i =[a i1 ,a i2 ] T Where T represents the transpose. The position of the target to be located is represented as x = [x1, x2]. T Anchor nodes can receive Received Signal Strength (RSS) information from targets, whether the signal is transmitted via radio or acoustic signals. This is expressed as an expression for the received signal strength:

[0034]

[0035] Among them, P rid0 represents the received transmission power value from the target at the i-th anchor node; d0 represents the reference distance value; P0 represents the target's transmission power; α represents the path loss factor, which is usually taken as [2,6]; ||·|| represents the second norm; γ i This indicates that the RSS ranging technique has a mean of zero and a variance of σ. i Gaussian distribution of wave-masking noise.

[0036] The received signal strength of the i-th anchor node is the transmitted power value from the target. The expression for the received signal strength is the ranging model.

[0037] S2. Select a reference node, perform a subtraction operation on the ranging model of the reference node, introduce a natural constant, construct a differential ranging model, and use a linear unbiased estimation method to solve for the initial position of the target.

[0038] The specific steps of S2 are as follows:

[0039] S21. Select a certain anchor node as the reference node, and let its received signal strength be P. r1 The corresponding ranging model can then be expressed as:

[0040]

[0041] S22. Subtract the ranging model corresponding to other anchor nodes from the ranging model in S21. If the prior information of the path loss factor is known, the target transmission power can be eliminated.

[0042]

[0043] S23. By changing the base of the logarithm in the ranging model that eliminates target transmission power in S22 and rearranging the terms, we can obtain:

[0044]

[0045] S24. By squaring the expression after changing the base in S23 and introducing the natural constant e, we get:

[0046]

[0047] S25. Using the first-order Taylor series expansion, the expression in S24 after introducing the natural constant e can be further transformed into:

[0048]

[0049] in,

[0050] S26. By rearranging terms in equation (6) of S25, we get:

[0051]

[0052] in,

[0053] S27. Let the target position θ = [x1, x2, χ] T , where χ=||x|| 2 Construct a differential ranging model, namely:

[0054]

[0055] in,

[0056] S28. Solve the differential ranging model using the linear unbiased estimation method to obtain the initial position of the target.

[0057] The initial position of the target is:

[0058]

[0059] in, ·|3 represents the first two terms of the vector; ·|3 represents the third term of the vector.

[0060] S3. Based on the initial target position, establish an objective optimization function with the path loss factor as the variable to be solved, construct constraints, and use the improved whale optimization algorithm to solve for the optimal path loss factor.

[0061] The specific steps for S3 are as follows:

[0062] S31. Initialize the target position Substituting into the equation in S26, we establish an objective optimization function with the path loss factor as the variable to be solved. The expression of the objective optimization function is:

[0063]

[0064] S32. To solve the objective function in S31, constraints are constructed based on the empirical value of the path loss factor. An improved whale optimization algorithm is introduced to solve the objective function through three steps: surrounding the prey, hunting behavior, and searching for the prey.

[0065] S33. In the prey encirclement phase, the update equation of the original whale optimization algorithm is improved by introducing perturbation and random numbers, namely:

[0066] Φ=|Cα * (t)-α(t)|,

[0067] α(t+1)=α * (t)-ΛΦ,

[0068] Where t represents the number of iterations; α * (t) represents the current optimal path loss factor; parameter C is the disturbance amount and C = 2Υ, where Υ represents a random number between 0 and 1; Λ is a random number with a value of [-1, 1].

[0069] S34. During the hunting behavior phase, the equations are updated by simulating the spiral hunting behavior of whales, i.e.:

[0070] α(t+1)=α * (t)+|α * (t)-α(t)|e bΛ cos(2πΛ),

[0071] Among them, e bΛ cos(2πΛ) is defined as the whale's spiral attack hunting tendency. However, since whale hunting behavior involves both contraction and spiral attack, and considering that the probabilities of contraction and spiral attack are equal at 50%, the update equation for hunting behavior can be further improved, namely:

[0072]

[0073] S35. The update equation for the prey-hunting phase is constructed as follows:

[0074] Φ=|Cα rand -α(t)|,

[0075] α(t+1)=α rand -ΛΦ,

[0076] Where, α rand It is a randomly selected whale position vector.

[0077] The specific steps for solving the optimal path loss factor using the improved whale optimization algorithm include:

[0078] Step 1: Initialize the population size and maximum number of iterations t max and related parameters;

[0079] The second step is to calculate the initial path loss factor based on the fitness optimization function.

[0080] The third step is to determine whether the probability p is greater than 0.5. If it is, proceed to the hunting behavior update equation; otherwise, determine whether the random number Λ is greater than 0.5. If it is, proceed to the surrounding prey behavior update equation; otherwise, proceed to the prey search update equation stage.

[0081] Step 4: Obtain the optimal path loss factor for the current position, and then determine whether the current iteration count is greater than t. maxIf so, output the optimal path loss factor; otherwise, iterate for t+1 and proceed to the second step to loop until the iteration count is greater than t. max Output the optimal path loss factor The flowchart of the improved whale optimization algorithm solution is as follows: Figure 2 As shown.

[0082] S4. Based on the optimal path loss factor, the differential ranging model is reconstructed, and the reconstructed differential ranging model is solved by the linear least squares method to obtain the optimized target position.

[0083] The specific steps of S4 are as follows:

[0084] S41, Optimal path loss factor Substituting into the ranging model in S22, we get:

[0085]

[0086] S42. After performing transposition and logarithmic operations, the reconstructed differential ranging model can be obtained, i.e.

[0087]

[0088] in,

[0089] S43. By using the linear least squares method to solve the reconstructed differential ranging model, the optimized target position can be obtained, i.e.:

[0090]

[0091] The optimized target location is the location of the required marine sensor network.

[0092] To verify the effectiveness of the Unknown Parameters Localization Algorithm (UPLA) for ocean sensor network localization provided by this invention, simulation experiments were conducted in Matlab R2021b. The dynamics of ocean height were simulated using a random walk model. Different localization methods were compared: Unconstrained Best Linear Unbiased Estimation (Ublue), Ratio and Search (RAS), and Accurate and Simple Source Localization (ASSL). The root mean square error (RMSE) was used as the evaluation criterion. The expression for RMSE is:

[0093]

[0094] Where x represents the actual location; The estimated position is indicated by MC; the total number of Monte Carlo simulations is 1000 in the simulation; and mc represents the current number of Monte Carlo simulations.

[0095] Other relevant fixed parameters are set as follows: the side length of the simulation area is 20m, α min =2, α max =6.

[0096] Figure 3 This chart compares the positioning errors of various methods under different noise conditions with N=10. As can be seen from the chart, the positioning performance of all methods decreases and the positioning error increases with increasing noise. When the noise level is low, the positioning errors of the UPLA method provided by this invention are similar to those of Ublue, exhibiting similar positioning performance. However, as the noise level increases, the superior positioning performance of UPLA becomes increasingly apparent, with a positioning error smaller than that of Ublue and higher accuracy than ASSL and RAS. Therefore, UPLA demonstrates better positioning performance under different noise environments and with unknown parameters.

[0097] Figure 4 For different numbers of anchor nodes, the algorithms in σ i 2 A comparison chart of positioning errors of 3dB is presented. The chart shows that the positioning accuracy of each algorithm improves with the increase in the number of anchor nodes. When the number of anchor nodes is small, the positioning errors of each algorithm are similar, and Ublue performs better than UPLA. However, as the number of anchor nodes increases, UPLA's positioning performance gradually surpasses Ublue's, and it has a smaller positioning error compared to ASSL and RAS. Therefore, when there is a certain number of anchor nodes, the UPLA method provided by this invention can achieve good positioning accuracy under conditions of parameter uncertainty.

[0098] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A method for locating marine sensor networks with uncertain parameters, characterized in that, Includes the following steps: S1. Construct a ranging model based on the received signal strength according to the signal propagation loss; S2. Select a reference node, perform a subtraction operation on the ranging model of the reference node, introduce the natural constant, construct a differential ranging model, and use the linear unbiased estimation method to solve for the initial position of the target. S3. Based on the initial target position, establish an objective optimization function with the path loss factor as the variable to be solved, construct constraints, and use the improved whale optimization algorithm to solve for the optimal path loss factor. S4. Based on the optimal path loss factor, the differential ranging model is reconstructed, and the reconstructed differential ranging model is solved by the linear least squares method to obtain the optimized target position. The specific steps of S2 are as follows: S21. Select a certain anchor node as the reference node, and let its received signal strength be... The corresponding ranging model can then be expressed as: S22. Perform the difference operation between the ranging model corresponding to other anchor nodes and the ranging model in S21: S23. Change the base of the logarithm in the ranging model that eliminates target transmission power in S22, and rearrange the terms to obtain: S24. Squaring the expression after changing the base in S23 and introducing the natural constant. Then, we get: S25. Further, the natural constant will be introduced through the first-order Taylor series expansion. The expression in S24 is then transformed into: in, ; ; S26. By rearranging terms in expression S25, we get: in, S27, Set the target position ,in Construct a differential ranging model, namely: in, S28. Solve the differential ranging model using the linear unbiased estimation method to obtain the initial position of the target; The initial position of the target is: in, ; ; ; Represents the first two terms of the vector; This represents the third term of the vector.

2. The method for locating marine sensor networks with uncertain parameters according to claim 1, characterized in that, The ranging model is as follows: in, Indicates the first The received signal strength of each anchor node; Indicates a reference distance value; Indicates the target's transmission power; Indicates the path loss factor; This represents Gaussian distributed wave-masking noise.

3. The method for locating marine sensor networks with uncertain parameters according to claim 2, characterized in that, Path loss factor .

4. The method for locating marine sensor networks with uncertain parameters according to claim 1, characterized in that, The improved whale optimization algorithm solves the objective optimization function through three steps: surrounding the prey, hunting behavior, and searching for the prey.

5. The method for locating marine sensor networks with uncertain parameters according to claim 4, characterized in that, The process of encircling the prey specifically involves introducing perturbations and random numbers to update the target optimization function.

6. The method for locating marine sensor networks with uncertain parameters according to claim 4, characterized in that, The hunting behavior is specifically a simulation of the spiral hunting behavior of whales, and the target optimization function is updated based on the spiral attack hunting trend of whales.

7. The method for locating marine sensor networks with uncertain parameters according to claim 6, characterized in that, The whale's spiral attack hunting trend includes contraction and spiral attack, both with a probability of 0.5.

Images