Relay position optimization method in double-hop multi-user multi-relay underwater wireless optical communication system

Abstract

The application discloses a relay position optimization method of a double-hop multi-user multi-relay underwater wireless optical communication system, and comprises the following steps: based on a certain system performance expression as a target function, using a Hungarian algorithm to allocate relay nodes for source nodes; based on the obtained source node and relay node pairing combination, using a particle swarm algorithm to optimize and update the positions of the selected relay nodes; based on the updated relay node positions, using the Hungarian algorithm again to allocate relay nodes for the source nodes, if the new allocation is different from the original allocation, continuing to use the particle swarm algorithm to optimize the positions of the selected relays until the current allocation is the same as the last allocation. The method considers the case that there may be a better allocation after single optimization of the relay position, can effectively improve the overall performance of the system, and can provide certain guiding significance for the relay deployment of the actual UWOC.

Description

Technical Field

[0001] This invention belongs to the field of underwater wireless optical communication technology. Specifically, it is an iterative optimization method that combines theoretical analysis and simulation research in a dual-hop multi-user multi-relay underwater wireless optical communication system, which jointly considers relay allocation and relay location, uses the Hungarian algorithm for relay allocation, and uses the particle swarm optimization algorithm to optimize relay location based on the system's average outage probability as a cost function. Background Technology

[0002] Underwater Wireless Optical Communication (UWOC) technology is a communication technology that uses light as a carrier and water as a propagation medium. Compared with traditional acoustic communication technology, it has the advantages of high transmission rate, low latency and low power consumption. The UWOC system is a high-efficiency communication solution that uses UWOC technology as a communication means and is suitable for a variety of underwater applications, including offshore oil field exploration, marine exploration, marine environmental monitoring, disaster prevention and control and port security.

[0003] However, the stability of UWOC systems is susceptible to light scattering and absorption effects and underwater optical turbulence (UOT). Particles in the water cause light scattering and absorption, resulting in optical power loss. UOT caused by bubbles, salinity, and temperature changes can easily cause optical link interruptions. Recent studies on UWOC channel statistical models have shown that the severity of light scattering and absorption effects and UOT impacts are related to the communication distance, and the impact increases with the distance. Moreover, using relays to transform long point-to-point UWOC systems from source node to destination node into short-link UWOC systems from source node to relay to destination node can reduce the scattering and absorption effects and UOT impacts on the optical link, improve the stability of the UWOC system, enhance communication quality, and reduce the overall system energy consumption.

[0004] Current research on UWOC systems largely focuses on constructing channel statistical models and analyzing performance, such as log-normal distribution models, mixed exponential-gamma distribution models, mixed exponential-generalized gamma distribution models, bit error rate, and channel capacity. Research on the deployment and optimization of relay locations in UWOC systems is largely unreported. Therefore, studying the optimization of relay locations in UWOC systems based on a certain cost function is a crucial research direction with significant practical implications for actual relay deployment. Summary of the Invention

[0005] An object of the present invention is to use the average outage probability of the links of an underwater wireless optical communication (UWOC) system as a measurement indicator of system performance, adopt the Hungarian algorithm to perform pairing selection of source nodes and relay nodes, perform position optimization for the selected relay nodes based on a particle swarm optimization algorithm, and perform alternating iteration to find the relay nodes that optimize system performance and the optimal positions thereof.

[0006] The present invention is implemented through the following technical solutions.

[0007] A relay position optimization method for a dual-hop multi-user multi-relay UWOC system, wherein the UWOC system comprises M source nodes S i , N relay nodes R j and one destination node D, which satisfies a redundant relay configuration of M<N; the relays adopt decode-and-forward (DF) and intensity modulation / direct detection (IM / DD) for transmission; underwater absorption and scattering path loss and lognormal weak optical turbulence are considered simultaneously, the average outage probability of the system is taken as an optimization cost function, and the method comprises the following steps:

[0008] 1) Constructing a network model of a three-dimensional dual-hop multi-user multi-relay UWOC system, setting underwater optical channel parameters, and deriving an expression of the average outage probability of the dual-hop DF relay link system;

[0009] 2) Obtaining an outage probability matrix according to links that can be formed by all source nodes and all relay nodes, constructing an M×N outage probability matrix, and with a goal of minimizing the average outage probability of system links, using the Hungarian algorithm to perform squaring solving on the matrix, allocate relay nodes for the source nodes to obtain an optimal allocation matrix of source nodes and relay nodes;

[0010] 3) Locking the selected nodes according to the allocation matrix, using a three-dimensional particle swarm optimization algorithm to optimize the position coordinates (x, y, z) of the selected relay nodes with link outage probability as fitness, obtaining optimized relay node positions and the optimal outage probability of the currently selected links, and updating the relay node position matrix and the corresponding link outage probability matrix;

[0011] 4) Repeating steps 2) to 3) until the current relay node allocation is the same as the previous relay node allocation.

[0012] In the above technical solution, the present invention further has a further limited solution as follows:

[0013] Further, in the step 1), establishing the dual-hop multi-user multi-relay UWOC system model is implemented through the following steps:

[0014] 1a) Providing a channel gain model considering the combined influence of turbulence and path loss:

[0015] H=αl (1)

[0016] Where α represents the fading amplitude caused by underwater optical turbulence (UOT), and l represents the path loss caused by scattering and absorption effects.

[0017] 1b) Give the expressions for the probability density function f(α) of the fading amplitude α and the path loss l:

[0018] The probability density function f(α) of the fading amplitude α is expressed as follows:

[0019]

[0020] The turbulence is modeled using a weakly turbulent distribution with a log-normal distribution, where α = exp(2X), and X is a variable satisfying the condition u. X The variance is Gaussian distribution, With flicker index The relationship is

[0021] The path loss l can be expressed as:

[0022]

[0023] Among them, A r η represents the aperture area of ​​the receiver. r and η t Let θ and θ0 represent the optical efficiency of the receiver and transmitter, respectively; let c(λ) represent the tilt angle and divergence angle of the beam, respectively; let θ0 represent the attenuation coefficient related to the wavelength λ of the light wave; let c(λ) represent the sum of the absorption coefficient and the scattering coefficient; and let d represent the transmission distance.

[0024] 1c) Establish a signal transmission model for a dual-hop multi-user multi-relay (UWOC) system:

[0025] The system adopts a network structure with multiple source nodes, multiple relay nodes, and one destination node. Source node S i To relay node R j The signal transmission process can be represented as:

[0026]

[0027] in, For relay node R j The received signal; ρ is the photoelectric response coefficient; For source node S i The transmission power; For source node S i To relay node R j Channel gain; For source node S i The signal sent; For relay node Rj The mean at point is 0, and the variance is . Additive white Gaussian noise.

[0028] Relay node R j The signal transmission process to destination node D can be represented as:

[0029]

[0030] in, The received signal for destination node D. For relay R j The transmission power; For relay node R j Channel gain to destination node D; For relay node R j The transmitted signal, For relay node R j The noise at the receiving end of the link to destination node D has a mean of 0 and a variance of . The Gaussian distribution.

[0031] Additive white Gaussian noise includes background noise, dark current noise, thermal noise, and shot noise. Background noise includes solar radiation noise and blackbody radiation noise.

[0032] Furthermore, in step 1), the expression for the average interruption probability of the system is derived through the following steps:

[0033] The average interruption probability expression for the UWOC two-hop link is derived from the average signal-to-noise ratio (SNR) of the link:

[0034] The SNR decision threshold is set to γ ​​when an interrupt occurs. th The probability of link interruption can be expressed as:

[0035]

[0036] in, Indicates from source node S i via relay node R j The received signal-to-noise ratio of the entire link to the destination node D. Indicates from source node S i Reaching relay node R j The link signal-to-noise ratio, Indicates from relay node R j The signal-to-noise ratio of the link reaching the destination node D.

[0037] After algebraic transformations and using relevant properties, the expression for the interruption probability of the UWOC double-hop link is obtained as follows:

[0038]

[0039] in, Indicates from source node S i Reaching relay node R j Path loss, Indicates from relay node R j The path loss to reach the destination node D. and These represent the origin from the source node S. i Reaching relay node R j The mean and standard deviation of the logarithmic fading amplitude suffered by the link. and These represent the relay node R. j The mean and standard deviation of the logarithmic fading amplitude experienced by the link to destination node D.

[0040] Furthermore, in step 2), the interruption probability matrix is ​​calculated through the following steps:

[0041] 2a) Calculate all source node S based on the location coordinates of the source node, relay node, and destination node. i To relay node R j and relay node R j Distance to destination node D and Calculate the noise variance at all relay nodes and destination nodes, calculate the scintillation index from all source nodes to relay nodes and from relay nodes to destination nodes, then calculate the mean and variance of the fading amplitude α from the scintillation index, and calculate the path loss l of all combined links.

[0042] 2b) Based on the interruption probability expression, calculate the interruption probability of the link consisting of all source nodes and relay nodes. For a UWOC system with M source nodes, N relay nodes, and one source node, an M×N interruption probability matrix P is obtained, where the matrix elements P... ij Indicates source node S i After passing through relay node R j The probability of the link to the destination node D being interrupted.

[0043] Furthermore, in step 2), the Hungarian algorithm is used to optimally allocate relay nodes to the source nodes, and the allocation matrix is ​​obtained through the following steps:

[0044] 2c) The interruption probability matrix is ​​expanded into a square matrix, which is done using the following method:

[0045] First, find the maximum value f in the interruption probability matrix of dimension M×N. max Then use a value that is all f maxThe (NM)×N matrix is ​​concatenated after the M×N interruption probability matrix, thus expanding the interruption probability matrix into an N×N square matrix.

[0046] 2d) Based on the expanded interruption probability matrix, the Hungarian algorithm is used to obtain the allocation matrix that minimizes the average interruption probability of the system. The first M×N elements of the allocation matrix are then taken to obtain the allocation matrix of the system.

[0047] Furthermore, in step 3), the particle swarm optimization algorithm is used to optimize the allocation of relay node positions and update the relay position and interruption probability matrix through the following steps:

[0048] 3a) Initialize the parameters of the particle swarm optimization algorithm, including the maximum number of iterations, population size, dimension, inertia weight, learning factor, velocity boundary, and position boundary;

[0049] 3b) Initialize the particle velocities and positions of the population based on the population size, dimension, velocity boundary, and position boundary;

[0050] 3c) Based on the initial particle and interruption probability expression, calculate the cost value of each particle, and initialize the global optimal position, global optimal cost value, individual optimal position, and individual optimal cost value;

[0051] 3d) In each iteration, the velocity of each particle in the population is updated according to the velocity update formula of the particle algorithm and the velocity is processed by boundary. Then the population particle position is updated and the population particle is processed by boundary. The value of the generation is updated according to the new population particle position. The global optimal position, global optimal value of generation, individual optimal position, and individual optimal value of generation are updated using the new value of generation.

[0052] 3e) Repeat step 3d) until the number of iterations is reached, and then update the global optimal position and global optimal cost for the selected relay node's position and the corresponding interruption probability matrix elements, respectively.

[0053] The present invention has the following advantages:

[0054] This invention first proposes an optimization method for relay positions in a two-hop multi-user multi-relay (UWOC) system, using the Hungarian algorithm for relay allocation and the particle swarm optimization algorithm for relay position optimization. This method solves the problem that optimizing relay positions based on a single relay allocation may not achieve the optimal configuration of relay nodes in the system. By iteratively combining the Hungarian algorithm and the particle swarm optimization algorithm, the system can achieve optimal performance. Furthermore, this method has practical guiding significance for the deployment of relay nodes in the UWOC system. Attached Figure Description

[0055] Figure 1aThis is a schematic diagram of the initial state of a dual-hop multi-user multi-relay UWOC system;

[0056] Figure 1b This is a schematic diagram of relay allocation and relay location optimization in a dual-hop multi-user multi-relay UWOC system;

[0057] Figure 2 This is the flowchart of the Hungarian algorithm;

[0058] Figure 3 This is the flowchart of the particle swarm optimization algorithm;

[0059] Figure 4a The simulation results are for a UWOC system with 3 source nodes and 5 relay nodes before and after performance optimization.

[0060] Figure 4b The simulation results show the relay position optimization in a UWOC system with 3 source nodes and 5 relay nodes.

[0061] Figure 5 It is an optimization simulation of the UWOC system performance under 3 source nodes and different numbers of relay nodes; Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of this invention clearer, the following detailed description is provided in conjunction with the accompanying drawings and specific embodiments. This embodiment is merely a illustrative description of the principles of this invention and does not represent any limitation thereof.

[0063] This invention discloses a method for optimizing relay positions in a two-hop multi-user multi-relay (UWOC) system, which uses the Hungarian algorithm for relay allocation and the particle swarm optimization algorithm for relay position optimization. The method is characterized by comprising the following steps:

[0064] Step 1), establishing a dual-hop multi-user multi-relay (UWOC) system model is achieved through the following steps:

[0065] 1a) Considering the effects of underwater optical turbulence (UOT) and water scattering and absorption on the UWOC system, a channel model considering the combined effects of turbulence and path loss is given:

[0066] H=αl (1)

[0067] Where α represents the attenuation amplitude caused by underwater optical turbulence, used to describe the effect of underwater optical turbulence; l represents the path loss, used to describe the effect of water scattering and absorption.

[0068] 1b) Give the expressions for the probability density function f(α) of the fading amplitude α and the path loss l:

[0069] The probability density function f(α) of the fading amplitude α is expressed as follows:

[0070]

[0071] The turbulence is modeled using a weakly turbulent distribution with a log-normal distribution, where α = exp(2X), and X is a variable satisfying the condition u. X The variance is The distribution is Gaussian. Here, α is normalized, i.e. get With flicker index satisfy The flicker index can be expressed as Where d represents the transmission distance, λ represents the wavelength of the light wave, κ is the scalar spatial frequency, and X(κ) represents the power spectrum affected by turbulence. Indicates the type of wavefront of light. Time represents a spherical wave. Time represents a plane wave.

[0072] The path loss l can be expressed as:

[0073]

[0074] Among them, A r η represents the aperture area of ​​the receiver. r and η t Let θ and θ0 represent the optical efficiencies of the receiver and transmitter, respectively, and let c(λ) represent the tilt angle and divergence angle of the beam, respectively. c(λ) represents the attenuation coefficient, which is the sum of the absorption coefficient and the divergence coefficient.

[0075] 1c) Establish a signal transmission model for a dual-hop multi-user multi-relay (UWOC) system:

[0076] The system adopts a network structure with multiple source nodes, multiple relay nodes, and one destination node, such as... Figure 1a A schematic diagram of a multi-user multi-relay UWOC system network structure with 3 source nodes, 5 relay nodes and 1 destination node is given. Figure 1b A schematic diagram illustrating the position changes of relay nodes in the system after optimization using the method of this invention is provided, where dashed lines represent the original positions and solid lines represent the optimized positions. For lack of generality, assume the system has M source nodes S. i There are N relay nodes R, i = 1, 2, ..., M. j j = 1, 2, ..., N, and M < N, source node S i To relay node R j The link is the first hop, relay node R j The link to destination node D is the second hop. The first hop is the source node S. i To relay node R j The signal transmission process can be represented as:

[0077]

[0078] in, For relay node R j The signal received is from the source node Si; ρ is the photoelectric response coefficient. For source node S i The transmission power; For source node S i To relay node R j Channel gain; For source node S i The signal sent; For relay node R j The mean at point is 0, and the variance is . Additive white Gaussian noise.

[0079] Assuming the relay node uses intensity modulation / direct detection (IM / DD) technology and adopts a decode-forward (DF) scheme, i.e., relay node R... j Signals from the sending node Restored to signal After being sent out, the second hop is the relay node R. j The signal transmission process to destination node D can be represented as:

[0080]

[0081] in, The received signal for destination node D. For relay node R j The transmission power; For relay node R j Channel gain to destination node D; For relay node R j The noise at the receiving end of the link to destination node D has a mean of 0 and a variance of . The Gaussian distribution.

[0082] The system considers noise including background noise, dark current noise, thermal noise, and shot noise. Background noise includes solar radiation noise and blackbody radiation noise. These noises can all be considered as additive white Gaussian noise. The noise variance of the background radiation can be calculated as follows: Where q represents the electron charge, ρ represents the response coefficient of the photodetector, and P BG This represents the background noise power, and P BG =P BG_solar +P BG_blackboby P BG_solar and P BG_blackbody Representing solar radiation power and blackbody radiation power respectively, respectively through and Calculation, where D represents the aperture diameter, θ FOV T represents the field of view, Δλ represents the optical filter bandwidth, and T represents the field of view angle. F E represents transmittance, R represents reflectance under downward spectral illuminance, and L represents the reflectance under downward spectral illuminance. fac Indicates the direct dependence on underwater radiation, d w Let represent depth, h represent Planck's constant, c represent the underwater speed of light, a represent the radiation absorption factor, T represent the Gaussian beam wavenumber, and B represent the electron bandwidth; the dark current noise variance can be calculated as... Among them, I dc Represents dark current; thermal noise variance can be calculated as... Among them, T e R represents the equivalent temperature, F represents the system noise figure, and R represents the system noise level. L Represents the load resistance; finally, the shot noise variance can be calculated as... Among them, P s Indicates signal power.

[0083] Furthermore, in step 1), the expression for the average interruption probability of the system is derived through the following steps:

[0084] 1d) The expression for the average outage probability of a two-hop UWOC link is derived from the average signal-to-noise ratio (SNR) of the link:

[0085] The average SNR is equal to the average power density of the signal divided by the average power density of the noise. Therefore, the average SNR of the first hop and the second hop can be expressed as:

[0086]

[0087] And because the relay uses DF technology, the source node S i The SNR to the destination node D depends on the minimum of the two hops, i.e. Let γ be the threshold for determining SNR when an interrupt occurs. th If the SNR is less than the decision threshold, the link is interrupted. In a two-hop DF trunk link, when one hop is interrupted, the entire link is interrupted; only when neither hop is interrupted will the link remain uninterrupted. Based on the properties of the cumulative distribution function, the link interruption probability can be expressed as:

[0088]

[0089] in, Indicates from source node S i via relay node R j The received signal-to-noise ratio of the entire link to the destination node D. Indicates from source node Si Reaching relay node R j The link signal-to-noise ratio, Indicates from relay node R j The signal-to-noise ratio of the link reaching the destination node D.

[0090] Substituting equations (1) and (6) into equation (7) and performing an algebraic transformation, we can obtain the function of the interruption probability with respect to α, expressed as:

[0091]

[0092] From step 1b), we know that α follows a log-normal distribution. Normalizing α yields a standard normal distribution. Then, based on the properties of the cumulative distribution function of the standard normal distribution, we can obtain the expression for the average interruption probability of the UWOC dual-hop link, as follows:

[0093]

[0094] Where Q(·) represents the Gaussian-Q function, Indicates from source node S i Reaching relay node R j Path loss, Indicates from relay node R j The path loss to reach the destination node D. and These represent the origin from the source node S. i Reaching relay node R j The mean and standard deviation of the logarithmic fading amplitude suffered by the link. and These represent the relay node R. j The mean and standard deviation of the logarithmic fading amplitude experienced by the link to destination node D.

[0095] Step 2), calculating the interruption probability matrix and pairing source nodes and relay nodes is achieved through the following steps:

[0096] 2a) Set the position coordinates of the source node, initial relay node, and destination node to a Cartesian coordinate system (x, y, z), where x and y are the coordinate axes in the horizontal plane, and the z-axis represents depth. Set the position coordinates of the source node, destination node, and initial relay node respectively. Calculate all source node S values ​​based on the position coordinates of the source node, relay node, and destination node. i To relay node R j and relay node R j Distance to destination node D and Calculate the noise variance at all relay nodes and destination nodes according to the noise variance calculation method in step 1c), calculate the flicker index from all source nodes to relay nodes and from relay nodes to destination nodes according to the flicker index and path loss l calculation method in step 1b), and then calculate the mean and variance of the fading amplitude α from the flicker index, and calculate the path loss l of the possible links.

[0097] 2b) According to equation (9), calculate the interruption probability of the link composed of all source nodes and relay nodes. For a UWOC system with M source nodes, N relay nodes, and one destination node, an M×N interruption probability matrix P is obtained, where the matrix elements P ij Indicates source node S i After passing through relay node R j The probability of the link to the destination node D being interrupted.

[0098] Furthermore, in step 2), the Hungarian algorithm is used to optimally allocate relay nodes to the source nodes, and the allocation matrix is ​​obtained through the following steps:

[0099] 2c) Since the relay nodes in the system are redundant, and the number of source nodes M is always less than the number of relay nodes N, the interruption probability matrix needs to be expanded into a square matrix. This can be done using the following method:

[0100] First, find the maximum value f in the interruption probability matrix of dimension M×N. max Then use a value that is all f max The (NM)×N matrix is ​​concatenated after the M×N interruption probability matrix, thus expanding the interruption probability matrix into an N×N square matrix.

[0101] 2d) Based on the expanded interruption probability matrix, the Hungarian algorithm is used to obtain the allocation matrix that minimizes the average interruption probability of the system. The first M×N elements of the allocation matrix are then taken to obtain the system's allocation matrix. The flowchart of the Hungarian algorithm is shown below. Figure 2 As shown.

[0102] Step 3), such as Figure 3 The particle swarm optimization (PSO) algorithm flowchart illustrates how it optimizes the allocation of relay node positions and updates the relay position and interruption probability matrices through the following steps:

[0103] 3a) Initialize the parameters of the particle swarm optimization algorithm, namely the maximum number of iterations, population size, dimension, inertia weight w, learning factors c1 and c2, velocity boundary and position boundary. The dimension is set to 3-dimensional, corresponding to the rectangular coordinates (x, y, z) of the spatial position of the relay node, that is, each particle is equivalent to a relay node.

[0104] 3b) Initialize the particle velocity V0 and particle position Loc0 of the population based on the population size, dimension, velocity boundary and position boundary, with the subscript indicating the iteration number;

[0105] 3c) Calculate the cost value of each particle according to Loc0 and Equation (9), and initialize the global optimal position (gBestLoc), global optimal cost value, individual optimal position (pBestLoc) and individual optimal cost value.

[0106] 3d) In the (i+1)th iteration, update the velocity V for each particle in the population. i+1 And for V i+1 Perform boundary processing, and then use the Loc formula. i+1 =Loc i +V i+1 Update population particle position Loc i+1 And perform boundary processing on the population particles; then, based on Loc... i+1 The value of the product is updated by formula (9), and then the new value of the product is used to update gBestLoc, global optimal value of product, pBestLoc and individual optimal value of product.

[0107] In step 3d), the velocity update of the particle swarm optimization algorithm is achieved through the velocity update formula, expressed as: V i+1 =w×V i +c1×rand×(Loc-pBestLoc)+c2×rand×(Loc-gBestLoc), where rand represents a random number with a value between 0 and 1; boundary processing is achieved by resetting the value of the dimension that exceeds the boundary, specifically by resetting the value of the dimension to the boundary value of the dimension.

[0108] 3e) Repeat step 3d) until the set maximum number of iterations is reached, and then update the global optimal position and global optimal cost to the position of the selected relay node and the corresponding interruption probability matrix elements, respectively.

[0109] In step 4), steps 2)-3) are repeated until the current relay node allocation is the same as the previous relay node allocation. Further explanation follows:

[0110] If the current relay node allocation is the same as the previous relay node allocation, it means that after optimizing the location of the previously allocated relays, it is still the optimal solution for the average performance of the system. If it is not satisfied, it means that the current relay allocation scheme is better than the previous relay allocation and location optimization result, and the location of the newly allocated relays needs to be optimized again to make the system achieve optimal performance.

[0111] The correctness and advantages of this invention can be further illustrated by comparing the following theoretical results:

[0112] In the method of this invention, simulation verification is performed using MATLAB to prove the effectiveness of the method.

[0113] Simulation results:

[0114] In a system consisting of 3 source nodes, 5 relay nodes, and 1 destination node, the transmit power of the source nodes and relay nodes is equally distributed. The method of this invention is used to perform simulation optimization of relay node positions under a total power of 34 dBm. The simulation results are as follows: Figure 4a and Figure 4b , Figure 4a The average outage probability of the system is given after optimizing the location of the initial relay nodes and the selected relay nodes. Figure 4b Simulation results for optimizing the relay node locations in this system are presented. Furthermore, with other variables remaining constant, we simulated systems with varying numbers of relay nodes using the method of this invention, and the system performance before and after optimization is shown below. Figure 5 As shown, with different numbers of relays, after optimizing the relay locations based on the relay node allocation, the average outage probability of the system basically converges to the same level. This indicates that, for a fixed source node, the method of this invention can effectively allocate relay nodes to the source node and optimize the locations of the allocated relay nodes, enabling the system to achieve optimal performance. Furthermore, in... Figure 5 In the study, it can be observed that the more relay nodes there are, the smaller the difference between the average interruption probability of the initially allocated relay nodes and the average interruption probability of the optimized system becomes. This is because when the number of relay nodes in the system increases, the probability that some relay nodes are close to the optimal relay node position of a certain source node increases. At this time, good performance can be obtained by using the Hungarian algorithm to allocate relay nodes. However, the subsequent application of the particle swarm optimization algorithm can still improve the performance of the system.

[0115] This invention is not limited to the above embodiments. Based on the technical solutions disclosed in this invention, those skilled in the art can make some substitutions and modifications to some of the technical features without creative effort, and all such substitutions and modifications are within the protection scope of this invention.

Claims

1. A relay location optimization method for a dual-hop multi-user multi-relay underwater wireless optical communication (UWOC) system, characterized in that, The UWOC system contains M source nodes S i N relay nodes R j One destination node D, satisfying M < N redundant relay configuration; the relay adopts DF decoding and forwarding, IM / DD intensity modulation / direct detection transmission, while considering underwater absorption and scattering path loss and log-normal weak optical turbulence, and using the system average outage probability as the optimization cost function, the method includes the following steps: 1) Construct a three-dimensional network model of a dual-hop multi-user multi-relay (UWOC) system, set underwater optical channel parameters, and derive the expression for the average outage probability of the dual-hop DF relay link system; 2) Obtain the interruption probability matrix based on the links that can be formed by all source nodes and all relay nodes, construct an M×N interruption probability matrix, and use the Hungarian algorithm to solve the matrix in a square form with the goal of minimizing the average interruption probability of the system links. Then, allocate relay nodes to the source nodes to obtain the optimal allocation matrix of source nodes and relay nodes. 3) Based on the allocation matrix, lock the selected node, use the three-dimensional particle swarm optimization algorithm with the link interruption probability as the fitness to optimize the position coordinates (x, y, z) of the selected relay node, obtain the optimized relay node position and the optimal interruption probability of the currently selected link, and update the relay node position matrix and the corresponding link interruption probability matrix. 4) Repeat steps 2)-3) until the current relay node allocation is the same as the previous relay node allocation.

2. The relay location optimization method for a dual-hop multi-user multi-relay underwater wireless optical communication system according to claim 1, characterized in that, In step 1), the establishment of the dual-hop multi-user multi-relay UWOC system model is achieved through the following steps: 1a) A channel gain model considering the combined effects of turbulence and path loss is given: H＝αl(1) Where α represents the fading amplitude caused by underwater optical turbulence (UOT); l represents the path loss caused by scattering and absorption effects; 1b) Give the expressions for the probability density function f(α) of the fading amplitude α and the path loss l: The probability density function f(α) of the fading amplitude α is expressed as follows: The turbulence is modeled using a weakly turbulent distribution with a log-normal distribution, where α = exp(2X), and X is a variable satisfying the condition u. X The variance is Gaussian distribution, With flicker index The relationship is The path loss l can be expressed as: Among them, A r η represents the aperture area of ​​the receiver. r and η t Let θ and θ0 represent the optical efficiency of the receiver and transmitter, respectively; let c(λ) represent the tilt angle and divergence angle of the beam, respectively; let θ0 represent the attenuation coefficient related to the wavelength λ of the light wave; let θ0 represent the sum of the absorption coefficient and the scattering coefficient; and let d represent the transmission distance. 1c) Establish a signal transmission model for a dual-hop multi-user multi-relay (UWOC) system: The system adopts a network structure with multiple source nodes, multiple relay nodes, and one destination node. Source node S i To relay node R j The signal transmission process can be represented as: in, For relay node R j The received signal; ρ is the photoelectric response coefficient; For source node S i The transmission power; For source node S i To relay node R j Channel gain; For source node S i The signal sent; For relay node R j The mean at point is 0, and the variance is . Additive white Gaussian noise; Relay node R j The signal transmission process to destination node D can be represented as: in, The received signal for destination node D. For relay R j The transmission power; For relay node R j Channel gain to destination node D; For relay node R j The transmitted signal, For relay node R j The noise at the receiving end of the link to destination node D has a mean of 0 and a variance of . The Gaussian distribution.

3. The relay location optimization method for a dual-hop multi-user multi-relay underwater wireless optical communication system according to claim 2, characterized in that, In step 1), the expression for the average system interruption probability is derived through the following steps: 1d) The expression for the average outage probability of a two-hop UWOC link is derived from the average signal-to-noise ratio (SNR) of the link: The SNR decision threshold is set to γ ​​when an interrupt occurs. th The probability of link interruption can be expressed as: in, Indicates from source node S i via relay node R j The received signal-to-noise ratio of the entire link to the destination node D. Indicates from source node S i Reaching relay node R j The link signal-to-noise ratio, Indicates from relay node R j The signal-to-noise ratio of the link reaching destination node D; After algebraic transformations and the application of relevant properties, the expression for the interruption probability of a double-hop UWOC link is obtained as follows: Where Q(·) represents the Gaussian-Q function, Indicates from source node S i Reaching relay node R j Path loss, Indicates from relay node R j The path loss to reach the destination node D. and These represent the origin from the source node S. i Reaching relay node R j The mean and standard deviation of the logarithmic fading amplitude suffered by the link. and These represent the relay node R respectively. j The mean and standard deviation of the logarithmic fading amplitude experienced by the link to destination node D.

4. The relay location optimization method for a dual-hop multi-user multi-relay underwater wireless optical communication system according to claim 1, characterized in that, In step 2), the interruption probability matrix is ​​calculated through the following steps: 2a) Calculate all source node S based on the location coordinates of the source node, relay node, and destination node. i To relay node R j and relay node R j Distance to destination node D and Calculate the noise variance at all relay nodes and destination nodes, calculate the scintillation index from all source nodes to relay nodes and from relay nodes to destination nodes, then calculate the mean and variance of the fading amplitude α from the scintillation index, and calculate the path loss l of all combined links. 2b) Based on the interruption probability expression, calculate the interruption probability of the link consisting of all source nodes and relay nodes. For a UWOC system with M source nodes, N relay nodes, and one source node, an M×N interruption probability matrix P is obtained, where the matrix elements P... ij Indicates source node S i After passing through relay node R j The probability of the link to the destination node D being interrupted.

5. The relay location optimization method for a dual-hop multi-user multi-relay underwater wireless optical communication system according to claim 1, characterized in that, In step 2), the Hungarian algorithm is used to optimally allocate relay nodes to the source nodes, and the allocation matrix is ​​obtained through the following steps: 2c) The interruption probability matrix is ​​expanded into a square matrix, which is done using the following method: First, find the maximum value f in the interruption probability matrix of dimension M×N. max Then use a value that is all f max The (NM)×N matrix is ​​concatenated after the M×N interruption probability matrix, thus expanding the interruption probability matrix into an N×N square matrix. 2d) Based on the expanded interruption probability matrix, the Hungarian algorithm is used to obtain the allocation matrix that minimizes the average interruption probability of the system. The first M×N elements of the allocation matrix are then taken to obtain the allocation matrix of the system.

6. The relay location optimization method for a dual-hop multi-user multi-relay underwater wireless optical communication system according to claim 1, characterized in that, In step 3), the optimization of relay node allocation and the updating of relay position and interruption probability matrices using the three-dimensional particle swarm optimization algorithm are achieved through the following steps: 3a) Initialize the parameters of the particle swarm optimization algorithm, set the particle dimension to 3-dimensional, corresponding to the underwater position coordinates (x, y, z), initialize the maximum number of iterations, population size, inertia weight, learning factor, velocity boundary and position boundary respectively; 3b) Initialize the particle velocity and three-dimensional particle position of the population based on the population size, dimension, velocity boundary, and position boundary; 3c) Based on the initial particle and interruption probability expression, calculate the cost function for each particle, and initialize the global optimal position, global optimal cost, individual optimal position, and individual optimal cost; 3d) In each iteration, the velocity of each particle in the population is updated according to the velocity update formula of the particle algorithm, and the velocity vector is processed by boundary. Then the population particle position is updated and the population particle is processed by boundary. The value of the generation is updated according to the new population particle position, and the global optimal position, global optimal value of generation, individual optimal position, and individual optimal value of generation are updated using the new value of generation. 3e) Repeat step 3d) until the set maximum number of iterations is reached, and then update the three-dimensional position and corresponding interruption probability matrix elements of the selected relay node with the global optimal position and global optimal cost respectively.

Images