Channel parameter estimation method and system based on adaptive ant colony optimization algorithm

Abstract

The invention discloses a channel parameter estimation method based on an adaptive ant colony optimization algorithm. The channel parameter estimation method comprises the following steps: step 1, generating a channel impulse response of a snapshot; step 2, initializing channel parameters of all paths, including time delay, angle and complex amplitude of each path; step 3, determining search spaces corresponding to all path channel parameters based on the initialized channel parameters; and step 4, based on the corresponding search space, estimating the channel parameter of each path by using a self-adaptive ant colony optimization algorithm, iteratively calculating the fitness of each path, and retaining the time delay, angle and complex amplitude corresponding to the optimal fitness in the current ant colony. According to the method and the system disclosed by the invention, aiming at the problem of estimating the multipath channel parameters, the accuracy of estimating the multipath channel parameters is improved by utilizing the capability of globally searching the optimal solution through the ant colony optimization algorithm and through a method of continuously iteratively calculating the multipath channel parameters.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a channel parameter estimation method and system based on an adaptive ant colony optimization algorithm in this field. Background Art

[0002] The transmission environment of wireless signals is dynamically changing. Signals are prone to encounter various obstacles during long-distance transmission, such as buildings, pedestrians, trees, vehicles, etc. Therefore, there are multiple paths for signal transmission. The multipath effect will cause dynamic changes in channel gain, phase, delay and other parameters. Accurate estimation of these multipath channel parameters is conducive to improving positioning accuracy and communication quality.

[0003] The current mainstream positioning systems include the Global Positioning System (GPS), GLONASS, and BeiDou Satellite Navigation System (BDS). However, in environments such as high-rise buildings, canyons, and valleys, the signal will be obstructed, which affects the positioning accuracy of the positioning system. By estimating the delay as one of the channel parameters, a distance-based positioning method can be implemented. This has attracted great attention as an auxiliary method for mainstream positioning systems. The key issue is to accurately estimate the channel parameters. At present, there are many methods for estimating channel parameters including azimuth, elevation, complex amplitude, delay, Doppler frequency, etc., such as MUSIC and ESPRIT, which use the orthogonal characteristics between the signal subspace and the noise subspace to estimate the channel parameters by eigenvalue decomposition; the SVD algorithm decomposes a matrix into the product of three matrices to extract and analyze its inherent linear structure and characteristics; Space Alternating Generalized Expectation (SAGE) optimizes the overall estimate of the objective function by alternately maximizing the conditional expectation of each parameter in the multidimensional parameter space. The MUSIC and ESPRIT algorithms have high computational complexity in large-scale arrays and high signal-to-noise ratio situations. The SVD algorithm has a high time complexity in calculating singular value decomposition when facing large matrices. The above three are all limited to the array structure of the transmitting and receiving antennas. Although the SAGE algorithm is not limited to the array structure of the transmitting and receiving antennas, in the high-dimensional parameter space, due to the alternating optimization of multiple parameters, the convergence speed may be slow. At the same time, SAGE performs poorly in dense multipath scenarios and is easily affected by multipath, resulting in a decrease in estimation accuracy and affecting the positioning effect.

[0004] Therefore, there is an urgent need for an accurate and efficient multipath parameter estimation method to solve the above-mentioned positioning accuracy problem. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a channel parameter estimation method and system based on an adaptive ant colony optimization algorithm, which can accurately estimate the multipath channel, reduce positioning errors, and improve the estimation accuracy of channel parameters.

[0006] The present invention adopts the following technical solution:

[0007] A channel parameter estimation method based on an adaptive ant colony optimization algorithm, the improvement of which is that it comprises the following steps:

[0008] Step 1, generating a snapshot of the channel impulse response;

[0009] Step 2: Initialize the channel parameters of all paths, including the delay, angle, and complex amplitude of each path;

[0010] Step 3: Based on the initialized channel parameters, determine the search space corresponding to all path channel parameters;

[0011] Step 4: Based on the corresponding search space, the adaptive ant colony optimization algorithm is used to estimate the channel parameters of each path, and its fitness is calculated in each iteration, and the delay, angle and complex amplitude corresponding to the optimal fitness in the contemporary ant colony are retained;

[0012] Step 5: When the end condition of the adaptive ant colony optimization algorithm is met, the estimated parameters are returned, and the delay, angle and complex amplitude corresponding to the optimal fitness in all generations of ant colonies are the estimated channel parameters; otherwise, return to step 4.

[0013] Furthermore, the impulse response h(t,τ) of the wideband channel established in step 1 is expressed as:

[0014]

[0015] In the above formula, L(t) is the number of arrival paths propagated at time t, α l (t), β l (t), a l (t) v l (t), τ l (t) respectively represent the horizontal arrival angle, elevation arrival angle, amplitude, phase, Doppler frequency, and delay of the lth arrival path, l = 1, ... L, d(α l (t),β l (t)) represents the steering vector of the receiving antenna defined by the angle of arrival, and δ is the impulse function.

[0016] Furthermore, in step 4, based on the corresponding search space, the channel parameters of each path are estimated using an adaptive ant colony optimization algorithm, which specifically includes:

[0017] Step 41, initializing and setting the number of ants in the ant colony, the number of iterations, the pheromone evaporation coefficient, the transfer probability constant and the local search step length;

[0018] Step 42, randomly generating initial positions of ants in the ant colony;

[0019] Step 43, calculating the fitness function value of the ant colony and setting it as the initial pheromone;

[0020] Step 44, calculating the state transition probability;

[0021] Step 45, performing location update using state transition rules;

[0022] Step 46, determine whether the ant has moved and update the pheromone;

[0023] Step 47, retaining the solution corresponding to the optimal fitness function value in the contemporary ant colony.

[0024] Furthermore, in step 41, the fitness function f is:

[0025]

[0026] In the above formula, Z is the received signal, f u represents the frequency of the u-th frequency band, u=1,…U, for:

[0027]

[0028] In the above formula, H is the frequency domain representation of the channel impulse response h.

[0029] Furthermore, in step 43, the algorithm initially sets the fitness function value corresponding to the randomly generated ant colony as the initial pheromone, and uses the pheromone concentration as a guide for ant path selection during the ant search process. The higher the pheromone concentration, the higher the importance and superiority of the path, making the ant more likely to choose the path. The value of the fitness function is converted into the pheromone concentration, and the quality information of the solution provided by the fitness function is used to guide the ants to tend to choose high-quality paths during the search process.

[0030] Furthermore, in step 44, the state transition probability is determined by comparing the pheromone concentration difference between the current path and the optimal path. The state transition probability is:

[0031]

[0032] In the above formula, ω best is the concentration of the path with the highest pheromone concentration, ω m is the pheromone concentration of the path where the mth ant is located.

[0033] Furthermore, in step 45, the position update uses a hybrid optimization method that combines local search and global search strategies. When the state transition probability is less than the transition probability constant, a local search is performed; when the state transition probability is greater than the transition probability constant, a global search is performed to generate a new ant position, and the boundary conditions are processed using the boundary absorption method to confine the ant position within the value range.

[0034] Furthermore, in step 46, an adaptive fitness-based evaporation coefficient is used, and when the optimal value obtained by the algorithm has no significant improvement within a certain number of cycles:

[0035]

[0036] In the above formula, ρ(t) is the pheromone evaporation coefficient of generation t, ρ min is the minimum pheromone evaporation coefficient, ρ max is the maximum pheromone evaporation coefficient, k is the mediation parameter, and f cur is the current fitness value, f best is the optimal fitness value in the current ant colony.

[0037] A channel parameter estimation system based on an adaptive ant colony optimization algorithm, the improvement of which is that it includes an initialization module and an estimation module;

[0038] The initialization module initializes the adaptive ant colony optimization algorithm parameters based on the channel impulse response, including delay, angle and complex amplitude;

[0039] The estimation module is based on the search space of the multipath parameter estimation problem and uses the global search capability of the adaptive ant colony optimization algorithm to estimate the delay, angle, and complex amplitude of the channel multipath. It calculates its fitness each iteration, retains the position information corresponding to the optimal fitness, and returns the estimated parameters according to the end conditions of the adaptive ant colony optimization algorithm.

[0040] The beneficial effects of the present invention are:

[0041] The method and system disclosed in the present invention improve the accuracy of estimating multipath channel parameters by using the ability of ant colony optimization algorithm to globally search for the optimal solution for the problem of estimating multipath channel parameters and by continuously iterating the method of calculating multipath channel parameters. It not only effectively balances the breadth and depth of the search space, but also avoids the local optimal trap common in traditional methods by guiding the search direction, and finally achieves more accurate and reliable multipath channel parameter estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic flow diagram of the method of the present invention;

[0043] Figure 2It is a detailed schematic diagram of the method of the present invention;

[0044] Figure 3 It is a specific flow chart of a single-transmit and single-receive antenna;

[0045] Figure 4 It is the convergence graph of the adaptive ant colony optimization algorithm;

[0046] Figure 5 The diagram is a Cramér-Rao boundary diagram and a variation trend of the mean square error of the first path delay estimated by the method of the present invention along with the delay difference between the two paths. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0048] Embodiment 1, this embodiment discloses a channel parameter estimation method based on adaptive ant colony optimization algorithm, such as Figure 1 , as shown in 2, comprising the following steps:

[0049] Step 1, generating a snapshot of the channel impulse response;

[0050] The impulse response model h(t,τ) of the broadband channel established in step 1 is expressed as:

[0051]

[0052] In the above formula, L(t) is the number of arrival paths propagated at time t, α l (t), β l (t), a l (t) v l (t), τ l (t) respectively represent the horizontal arrival angle, elevation arrival angle, amplitude, phase, Doppler frequency, and delay of the lth arrival path, l = 1, ... L, d(α l (t),β l (t)) represents the steering vector of the receiving antenna defined by the angle of arrival, and δ is the impulse function.

[0053] The number of path estimates at time t is L, so the channel impulse response at time t is:

[0054]

[0055] Performing discrete Fourier transform (DFT) on the above equation gives the frequency response of the channel:

[0056]

[0057] The adaptive ant colony optimization algorithm is used to estimate the path channel parameters and calculate the fitness. The fitness function f is:

[0058]

[0059] In the above formula, Z is the received signal, f u represents the frequency of the u-th frequency band, u=1,…U, for:

[0060]

[0061] In the above formula, H is the frequency domain representation of the channel impulse response h.

[0062] Step 2: Initialize the channel parameters of all paths, including the delay, angle, and complex amplitude of each path;

[0063] Step 3: Based on the initialized channel parameters, determine the search space corresponding to all path channel parameters;

[0064] Step 4: Based on the corresponding search space, the adaptive ant colony optimization algorithm is used to estimate the channel parameters of each path, and its fitness is calculated in each iteration, and the delay, angle and complex amplitude corresponding to the optimal fitness in the contemporary ant colony are retained;

[0065] Based on the corresponding search space, the adaptive ant colony optimization algorithm is used to estimate the channel parameters of each path, including:

[0066] Step 41, initializing and setting the number of ants in the ant colony, the number of iterations, the pheromone evaporation coefficient, the transfer probability constant and the local search step length;

[0067] Step 42, randomly generating initial positions of ants in the ant colony;

[0068] Step 43, calculating the fitness function value of the ant colony and setting it as the initial pheromone;

[0069] Step 44, calculating the state transition probability;

[0070] Step 45, performing location update using state transition rules;

[0071] Step 46, determine whether the ant has moved and update the pheromone;

[0072] Step 47, retaining the solution corresponding to the optimal fitness function value in the contemporary ant colony.

[0073] Step 5: When the end condition of the adaptive ant colony optimization algorithm is met, the estimated parameters are returned, and the delay, angle and complex amplitude corresponding to the optimal fitness in all generations of ant colonies are the estimated channel parameters; otherwise, return to step 4.

[0074] After the ant colony performs path search, the solution corresponding to the optimal fitness function value in the current ant colony is retained, and the final output is the solution corresponding to the optimal fitness function value in all generations of ant colonies, which is the estimated channel parameter.

[0075] Take a single-transmit and single-receive antenna as an example. Figure 3 As shown, the specific process is as follows:

[0076] Step 1, simulate and generate a snapshot of the channel impulse response, the channel impulse response sampling number is 128;

[0077] Step 2: Initialize the channel parameters of all paths based on the serial interference cancellation method, including the delay and complex amplitude of each path;

[0078] Step 3: Based on the initialized channel parameters, determine the search space corresponding to all path channel parameters;

[0079] Step 4: Based on the corresponding search space, the adaptive ant colony optimization algorithm is used to start estimating the channel parameters of each path, including delay and complex amplitude. The specific steps are as follows:

[0080] Step 41, initializing and setting the number of ants in the ant colony, the number of iterations, the pheromone evaporation coefficient, the transfer probability constant and the local search step length;

[0081] The frequency response of the channel is:

[0082]

[0083] In the above formula, a l (t) v l (t), τ l (t) represent the horizontal arrival angle, elevation arrival angle, amplitude, phase, Doppler frequency and delay of the lth (l=1,…L) arrival path respectively.

[0084] The fitness function f is:

[0085]

[0086] In the above formula, Z is the received signal, f u represents the frequency u=1,…U of the u-th frequency band, for:

[0087]

[0088] Step 42, using random numbers to randomly generate initial positions of ants in the ant colony;

[0089] Step 43, at the beginning of the algorithm, the fitness function value corresponding to the randomly generated ant colony is set as the initial pheromone, and the pheromone concentration is used as a guide for ant path selection during the ant search process. The higher the pheromone concentration, the higher the importance and superiority of the path, making the ant more likely to choose the path. The value of the fitness function is converted into the pheromone concentration, and the quality information of the solution provided by the fitness function is used to guide the ants to tend to choose high-quality paths during the search process.

[0090] Step 44, setting of transition probability, specifically includes:

[0091] The state transition probability is determined by comparing the pheromone concentration difference between the current path and the optimal path. The state transition probability is:

[0092]

[0093] In the above formula, ω best is the concentration of the path with the highest pheromone concentration, ω m is the pheromone concentration of the path where the mth ant is located.

[0094] Step 45, position update uses a hybrid optimization method that combines local search and global search strategies. The main purpose is to effectively balance local development and global exploration during the search process to improve the search efficiency of the algorithm and avoid falling into local optimality. When the state transition probability is less than the transition probability constant, a local search is performed; when the state transition probability is greater than the transition probability constant, a global search is performed to generate a new ant position, and the boundary conditions are processed using the boundary absorption method to limit the ant position within the value range.

[0095] Step 46, determine whether the ant has moved and update the pheromone;

[0096] This process requires dynamic adjustment of the pheromone evaporation coefficient. When the problem scale is relatively large, the existence of the pheromone evaporation coefficient will reduce the amount of information that has never been searched to close to 0, and the global search ability of the algorithm will be greatly reduced; however, when the pheromone evaporation coefficient is too large and the amount of information of the solution increases, the possibility of the previously searched solution being selected again increases, affecting the global search ability of the algorithm; although the global search ability of the algorithm can be improved by reducing the pheromone evaporation coefficient, it will also reduce the convergence speed of the algorithm. Therefore, it is chosen to adaptively change the value of the pheromone evaporation coefficient ρ, specifically using an adaptive fitness-based evaporation coefficient. When the optimal value obtained by the algorithm has no obvious improvement within a certain number of cycles, ρ is reduced to:

[0097]

[0098] In the above formula, ρ(t) is the pheromone evaporation coefficient of generation t, ρmin is the minimum pheromone evaporation coefficient, ρ max is the maximum pheromone evaporation coefficient, k is the mediation parameter, and f cur is the current fitness value, f best is the optimal fitness value in the current ant colony.

[0099] Step 47, retaining the solution corresponding to the optimal fitness function value in the contemporary ant colony.

[0100] Step 5: When the end condition of the adaptive ant colony optimization algorithm is met, the estimated parameters are returned, and the delay, angle and complex amplitude corresponding to the optimal fitness in all generations of ant colonies are the estimated channel parameters; otherwise, return to step 4.

[0101] In order to verify the effectiveness of the method of the present invention, the application effect of the present invention is described in detail in combination with the simulation results. In the specific simulation, two paths are set, the real amplitude of the first path is set to 1, the real time delay τ1 is set to 1, the real amplitude of the second path is set to 2, the real time delay τ2 of the second path is set to Δτ+τ1, which changes with Δτ, and the phase difference between the two paths is set to 50 degrees.

[0102] Root mean square error of estimated delay N is the Monte Carlo number, which is set to 300 in the experiment, τ is the actual path delay, is the n-th estimated path delay.

[0103] Figure 4 This is the convergence diagram of the adaptive ant colony optimization algorithm. It can be seen that the adaptive ant colony optimization algorithm has strong convergence and can quickly find the optimal value within a certain number of generations.

[0104] Figure 5 The figure shows the variation trend of the mean square error of the first path delay estimated by the method of the present invention with the delay difference between the two paths and the Cramer-Rao boundary diagram. It can be seen from the figure that when the delay difference between the two paths is between 0 and 0.5, the mean square error of the first path delay increases with the increase of the delay difference between the two paths. When the delay difference between the two paths is greater than 0.5, the mean square error of the first path delay is constantly approaching the performance boundary of the delay, that is, the Cramer-Rao boundary.

[0105] This embodiment also discloses a channel parameter estimation system based on an adaptive ant colony optimization algorithm, including an initialization module and an estimation module;

[0106] The initialization module initializes the adaptive ant colony optimization algorithm parameters based on the channel impulse response, including delay, angle and complex amplitude;

[0107] The estimation module is based on the search space of the multipath parameter estimation problem and uses the global search capability of the adaptive ant colony optimization algorithm to estimate the delay, angle, and complex amplitude of the channel multipath. It calculates its fitness each iteration, retains the position information corresponding to the optimal fitness, and returns the estimated parameters according to the end conditions of the adaptive ant colony optimization algorithm.

Claims

1. A channel parameter estimation method based on an adaptive ant colony optimization algorithm, characterized in that: The steps include: Step 1, generating a snapshot of the channel impulse response; Step 2: Initialize the channel parameters of all paths, including the delay, angle, and complex amplitude of each path; Step 3: Based on the initialized channel parameters, determine the search space corresponding to all path channel parameters; Step 4: Based on the corresponding search space, the adaptive ant colony optimization algorithm is used to estimate the channel parameters of each path, and its fitness is calculated in each iteration, and the delay, angle and complex amplitude corresponding to the optimal fitness in the contemporary ant colony are retained; Step 5: When the end condition of the adaptive ant colony optimization algorithm is met, the estimated parameters are returned, and the delay, angle and complex amplitude corresponding to the optimal fitness in all generations of ant colonies are the estimated channel parameters; otherwise, return to step 4.

2. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 1, characterized in that: The impulse response h(t,τ) of the wideband channel established in step 1 is expressed as: In the above formula, L(t) is the number of arrival paths propagated at time t, α l (t), β l (t), a l (t) v l (t), τ l (t) respectively represent the horizontal arrival angle, elevation arrival angle, amplitude, phase, Doppler frequency, and delay of the lth arrival path, l = 1, ... L, d(α l (t),β l (t)) represents the steering vector of the receiving antenna defined by the angle of arrival, and δ is the impulse function.

3. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 1, characterized in that: In step 4, based on the corresponding search space, the channel parameters of each path are estimated using an adaptive ant colony optimization algorithm, which specifically includes: Step 41, initializing and setting the number of ants in the ant colony, the number of iterations, the pheromone evaporation coefficient, the transfer probability constant and the local search step length; Step 42, randomly generating initial positions of ants in the ant colony; Step 43, calculating the fitness function value of the ant colony and setting it as the initial pheromone; Step 44, calculating the state transition probability; Step 45, performing location update using state transition rules; Step 46, determine whether the ant has moved and update the pheromone; Step 47, retaining the solution corresponding to the optimal fitness function value in the contemporary ant colony.

4. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 3 is characterized in that: In step 41, the fitness function f is: In the above formula, Z is the received signal, f u represents the frequency of the u-th frequency band, u=1,…U, for: In the above formula, H is the frequency domain representation of the channel impulse response h.

5. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 3 is characterized in that: In step 43, the algorithm initially sets the fitness function value corresponding to the randomly generated ant colony as the initial pheromone, and uses the pheromone concentration as a guide for ant path selection during the ant search process. The higher the pheromone concentration, the higher the importance and superiority of the path, making the ant more likely to choose the path. The value of the fitness function is converted into the pheromone concentration, and the quality information of the solution provided by the fitness function is used to guide the ants to tend to choose high-quality paths during the search process.

6. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 3, characterized in that: In step 44, the state transition probability is determined by comparing the pheromone concentration difference between the current path and the optimal path. The state transition probability is: In the above formula, ω best is the concentration of the path with the highest pheromone concentration, ω m is the pheromone concentration of the path where the mth ant is located.

7. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 3 is characterized in that: In step 45, the position update uses a hybrid optimization method that combines local search and global search strategies. When the state transition probability is less than the transition probability constant, a local search is performed; when the state transition probability is greater than the transition probability constant, a global search is performed to generate a new ant position, and the boundary conditions are processed using the boundary absorption method to confine the ant position within the value range.

8. The channel parameter estimation method based on the adaptive ant colony optimization algorithm according to claim 3, characterized in that: In step 46, an adaptive fitness-based evaporation coefficient is used. When the optimal value obtained by the algorithm does not improve significantly within a certain number of cycles: In the above formula, ρ(t) is the pheromone evaporation coefficient of generation t, ρ min is the minimum pheromone evaporation coefficient, ρ max is the maximum pheromone evaporation coefficient, k is the mediation parameter, and f cur is the current fitness value, f best is the optimal fitness value in the current ant colony.

9. A channel parameter estimation system based on an adaptive ant colony optimization algorithm, characterized in that: Includes an initialization module and an estimation module; The initialization module initializes the adaptive ant colony optimization algorithm parameters based on the channel impulse response, including delay, angle and complex amplitude; The estimation module is based on the search space of the multipath parameter estimation problem and uses the global search capability of the adaptive ant colony optimization algorithm to estimate the delay, angle, and complex amplitude of the channel multipath. It calculates its fitness each iteration, retains the position information corresponding to the optimal fitness, and returns the estimated parameters according to the end conditions of the adaptive ant colony optimization algorithm.

Images