A sparse circular array interference suppression beamforming method based on combined particle swarm optimization and diagonal loading factor update
Through the combined method of combining particle swarm algorithm and diagonal loading factor update, the problem of insufficient suppression effect of diagonal loading beamforming under different interference situations and insufficient applicability to conformal sparse arrays is solved, and faster convergence speed is achieved and better interference suppression effects, reducing system complexity and cost.
Patent Information
- Application Number
- CN202411203478.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-08-30
AI Technical Summary
The existing diagonal loading beamforming method is difficult to efficiently and accurately select suitable diagonal loading factors under different interference situations, resulting in a reduced suppression effect of interference signals and is not suitable for conformal sparse arrays.
Using the method of combining particle swarm algorithm and diagonal loading factor update, the algorithm convergence speed is improved by limiting the selection range of diagonal loading factor, and the algorithm convergence speed is improved by using chaotic sequences and velocity weight updates, and the local optimal solution is jumped out of the cross-section and mutation strategies to achieve sparseness. Adaptive interference suppression beamforming of circular arrays.
The convergence speed and interference suppression effect of the algorithm are improved, the array element and system cost are reduced, and the effective suppression of interfering signals is achieved.
Smart Images

Figure CN118740224B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of communication, aims at the communication difficulty problem of remote schools in distance education, and designs a sparse circular array interference suppression beamforming method by combining a particle swarm algorithm with a diagonal loading factor update. Background Art
[0002] When addressing the communication difficulties of remote schools in distance education, interference suppression beamforming is widely used in the field of mobile communications with complex interference propagation environments by flexibly adjusting the directional pattern beam direction generated by the antenna array, with the main lobe pointing in the desired direction and the zero line pointing in the direction of the interference signal. In order to further improve the robustness and robustness of the algorithm, in the prior art, Carlson and other scholars proposed beamforming based on the idea of loading matrix diagonal elements, called diagonal loading beamforming (DL). The DL algorithm increases the robustness of beamforming by adding a diagonal noise matrix to the covariance matrix obtained by sampling. It is relatively simple to implement and does not require complex signal processing algorithms and a large amount of computing resources. It can reduce the complexity of the communication system to a certain extent. This is especially important for resource-constrained or power-sensitive application scenarios, such as various multi-antenna communication scenarios such as cellular networks, wireless LANs and satellite communications. The main problem with the DL method is that the diagonal loading empirical values are currently mostly used. It is difficult to efficiently and accurately select suitable diagonal loading factor values under different interference conditions, thereby reducing the algorithm's suppression effect on interference signals.
[0003] At present, some methods for calculating the diagonal loading factor have been proposed, mainly combining the optimization algorithm theory for calculation. Among them, the representative ones are the use of convex optimization algorithms such as the steepest descent method to calculate the diagonal loading factor, and the use of swarm intelligence optimization algorithm to calculate the diagonal loading factor based on the determined diagonal loading factor range; the classic convex optimization method is to transform the non-deterministic constraints of the array direction vector into second-order cone optimization, reduce the computational complexity of the algorithm and solve the optimal diagonal loading factor. There are also methods that use the iterative update idea in the DL algorithm to continuously generate new diagonal loading factors and then determine the optimal diagonal loading factor; however, the current interference Suppression beamforming is mainly studied theoretically for linear arrays, and there is a lack of research on adaptive beamforming for conformal sparse arrays. In fact, interference suppression adaptive beamforming based on conformal sparse arrays can reduce the algorithm calculation amount and utilize the shape advantages of conformal arrays under the condition of sparse arrays, thereby reducing the complexity and cost of installing antenna systems. However, the layout and algorithm complexity of conformal sparse arrays bring huge challenges to related research, resulting in the number and depth of existing research being slightly insufficient. For this reason, we propose a sparse circular array interference suppression beamforming method that combines a particle swarm algorithm with a diagonal loading factor update. Summary of the invention
[0004] The purpose of the present invention is to provide a sparse circular array interference suppression beamforming method which combines a combined particle swarm algorithm with a diagonal loading factor update. Aiming at the problems that the existing diagonal loading beamforming method cannot determine the optimal diagonal loading factor, the algorithm has a slow convergence speed and is not suitable for conformal sparse arrays, the present invention discloses a conformal sparse array interference suppression beamforming method which combines a diagonal loading factor update (UDL) with a combined particle swarm algorithm (CPSO); the method avoids the negative impact of improper diagonal loading factor value on beamforming by limiting the selection range of the diagonal loading factor, adopts a combined particle swarm algorithm to improve the convergence speed of the algorithm through chaotic sequence and speed weight update, and uses crossover and mutation strategies to solve the problem that the algorithm is prone to fall into a local optimum and cannot jump out.
[0005] The technical solution adopted by the present invention is as follows:
[0006] A sparse circular array interference suppression beamforming method combining a particle swarm algorithm and a diagonal loading factor update comprises the following steps:
[0007] Step 1: First, according to the signal-to-noise ratio SNR, sparse circular array flow pattern , expected direction Interference signal direction , get the array receiving signal model , expected received signal model Received signal model with interference signal : Step 2: Calculate the noise and interference covariance matrix: ; Calculated based on the interference and noise covariance matrix Take the value range and use it as the algorithm search space, ; Step 3: The sum of the sidelobe levels at the interference positions of the diagonally loaded beamforming is used as the evaluation function, ;
[0008] Step 4: Optimize using combined particle swarm algorithm:
[0009] The specific steps include:
[0010] Step 401: Initialize algorithm parameters and set fitness function;
[0011] Step 402: Generate an initial population using a chaotic sequence;
[0012] Step 403: Enter the algorithm loop and set the speed weight update strategy;
[0013] Step 404: whether the iteration is completed, if the iteration is completed, output the population;
[0014] Step 405: If the iteration is not completed, the fitness function is calculated, and the population is updated using a crossover mutation strategy to complete interference suppression beamforming.
[0015] Infinite folding iterative chaotic map (ICMIC) initializes the diagonal loading factor population;
[0016] Infinite folding iterative chaotic mapping is used instead of the traditional method of initially generating a uniformly distributed random population. Through chaotic mapping, a large number of non-repeating points located above and below the value range are generated. This method can achieve better boundary traversal search capabilities; its mathematical expression is as follows: in a is a constant distributed over within the range.
[0017] In order to improve the convergence speed of the particle swarm algorithm, we propose a speed weight update strategy.
[0018] ;in ;
[0019] in, w is called the inertia weight, w max , w min is the value range of inertia weight, c max , c min is the value range of acceleration weight, t is the current iteration number, T is the maximum number of iterations; the particle swarm speed will gradually decrease with the increase of the number of iterations. At the beginning of the algorithm, it will perform global exploration at a relatively high speed, and at the later stage, it will perform local search at a relatively low speed, which will speed up the calculation speed and improve the optimization ability.
[0020] Through the crossover and mutation operations of genes, genetic algorithms have the ability to jump out of local optimal solutions, powerful global search capabilities and parallel information processing capabilities; however, crossover and mutation operations lead to large computational complexity and slow convergence of genetic algorithms. Considering that particles in the particle swarm algorithm have memory, fast convergence speed but are difficult to jump out of local optimal solutions, a particle swarm individual crossover and mutation strategy is designed; this strategy helps the algorithm jump out of local optimal solutions, increases the flexibility of particle changes, overcomes the problem of insufficient search accuracy of the particle swarm algorithm, and realizes the complementary advantages of different algorithms;
[0021] After the particle swarm updates the individual extreme value and the overall extreme value process, the crossover mutation strategy is introduced; the genetic population is defined GAPOP , its population number is the same as the particle swarm population number N Consistent; define the number of population crossovers , population variation ;
[0022] Generate crossover population ;
[0023] in GA pc is the crossover population selected from the genetic population.
[0024] Generate mutant population ;
[0025] in, GA pm is the variant population selected from the genetic population, p max is the maximum value within the particle variation range, p min is the minimum value within the particle variation range, D It is the population dimension.
[0026] The final population update is ; Complete interference suppression beamforming.
[0027] The technical effects achieved by the present invention are:
[0028] In the present invention, the particle swarm algorithm is improved, an infinite folding iterative chaotic mapping is used to generate an initial population, a speed weight update is used to accelerate the convergence speed, a genetic mutation function is used to help the algorithm jump out of the local optimal solution, and in view of the problem that the optimal diagonal loading factor cannot be determined, the selection range of the diagonal loading factor is limited to avoid the negative impact of its inappropriate value on beamforming. Based on the maximum interference suppression degree as the criterion, an innovative design is proposed for the adaptive beamforming of a sparse circular array, which achieves the suppression effect of interference signals while reducing the number of array elements and reducing the system cost.
[0029] In the present invention, a novel conformal sparse array interference suppression beamforming method is designed, which can achieve interference suppression to a greater extent than traditional methods, and is the first time to design a beamforming method based on a conformal sparse array, which has the effect of reducing the number of array elements and reducing system costs.
[0030] In the present invention, the advantages of the combined particle swarm algorithm being faster and more efficient are utilized to improve the speed of iterative convergence and the degree of interference suppression.
[0031] In the present invention, in the improved method of combined particle swarm algorithm, infinite folding iterative chaotic mapping is used to initialize the population, generating a large number of non-repeated points located above and below the value range, and this method can achieve better boundary traversal search capability. A speed weight update strategy is also applied to improve the convergence speed of the particle swarm algorithm. A particle swarm individual crossover mutation strategy is designed. This strategy assists the algorithm to jump out of the local optimal solution, increases the flexibility of particle changes, overcomes the problem of insufficient search accuracy of the particle swarm algorithm, and realizes the complementary advantages of different algorithms. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 It is a flow chart of a sparse circular array interference suppression beamforming method combining a particle swarm algorithm and a diagonal loading factor update according to the present invention;
[0033] Figure 2 It is the flow chart of the combined particle swarm algorithm of the present invention;
[0034] Figure 3 It is a comparison chart of the convergence speed of the combined particle swarm algorithm of the present invention;
[0035] Figure 4 It is a comparison diagram of the beamforming effect of the diagonal loading factor update and the fixed diagonal loading factor of the present invention;
[0036] Figure 5 This is a schematic diagram of the layout of the sparse circular array elements of the present invention;
[0037] Figure 6 This is a comparison diagram of the beamforming effects of sparse and non-sparse arrays of the present invention;
[0038] Figure 7 It is a diagram of the sparse array beamforming effect under three interferences of the present invention;
[0039] Figure 8 This is a diagram of the sparse array beamforming effect under four interferences of the present invention;
[0040] Fig. 9 It is a diagram of the sparse array beamforming effect under five interferences of the present invention;
[0041] Fig.10 is a diagram of the beam scanning capability of the sparse circular array of the present invention;
[0042] Fig.11 It is a flow chart of the diagonal loading factor updating algorithm of the present invention. DETAILED DESCRIPTION
[0043] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is specifically described below in conjunction with embodiments. It should be understood that the following text is only used to describe one or several specific embodiments of the present invention, and does not strictly limit the scope of protection of the specific claims of the present invention.
[0044] like Figure 1 As shown, a sparse circular array interference suppression beamforming method combining a particle swarm optimization algorithm and a diagonal loading factor update comprises the following steps:
[0045] Step 1: First, according to the signal-to-noise ratio SNR, sparse circular array flow pattern , expected direction Interference signal direction , get the array receiving signal model , expected received signal model Received signal model with interference signal : Step 2: Calculate the noise and interference covariance matrix: ;
[0046] Calculated based on the interference and noise covariance matrix Take the value range and use it as the algorithm search space, ;
[0047] In the formula is the diagonal loading factor, Is to construct a diagonal matrix, specifically the covariance matrix Converted into a diagonal matrix, is to find the standard deviation of the diagonal matrix; , is to find the covariance matrix The matrix inverse of
[0048] Step 3: The sum of the sidelobe levels at the interference positions of the diagonally loaded beamforming is used as the evaluation function. ;
[0049] like Figure 2 As shown,
[0050] Step 4: Optimize using combined particle swarm algorithm:
[0051] The specific steps include:
[0052] Step 401: Initialize algorithm parameters and set fitness function;
[0053] Step 402: Generate an initial population using a chaotic sequence;
[0054] Step 403: Enter the algorithm loop and set the speed weight update strategy;
[0055] Step 404: whether the iteration is completed, if the iteration is completed, output the population;
[0056] Step 405: If the iteration is not completed, the fitness function is calculated, and the population is updated using a crossover mutation strategy to complete interference suppression beamforming.
[0057] Infinite folding iterative chaotic map (ICMIC) initializes the diagonal loading factor population;
[0058] Infinite folding iterative chaotic mapping is used instead of the traditional method of initially generating a uniformly distributed random population. Through chaotic mapping, a large number of non-repeating points located above and below the value range are generated. This method can achieve better boundary traversal search capabilities; its mathematical expression is as follows: in a is a constant distributed over In order to improve the convergence speed of the particle swarm algorithm, we propose a speed weight update strategy; ;
[0059] in in, w is called the inertia weight, w max , w min is the value range of inertia weight, c max , c min is the value range of acceleration weight, t is the current iteration number, T is the maximum number of iterations. The particle swarm speed will gradually decrease with the increase of the number of iterations. At the beginning of the algorithm, it will perform global exploration at a relatively high speed, and at the later stage, it will perform local search at a relatively low speed, which will speed up the calculation speed and improve the optimization ability.
[0060] Through the crossover and mutation operations of genes, genetic algorithms have the ability to jump out of local optimal solutions, powerful global search capabilities and parallel information processing capabilities; however, crossover and mutation operations lead to large computational complexity and slow convergence of genetic algorithms. Considering that particles in the particle swarm algorithm have memory, fast convergence speed but are difficult to jump out of local optimal solutions, a particle swarm individual crossover and mutation strategy is designed; this strategy helps the algorithm jump out of local optimal solutions, increases the flexibility of particle changes, overcomes the problem of insufficient search accuracy of the particle swarm algorithm, and realizes the complementary advantages of different algorithms;
[0061] like Fig.11 As shown in the figure, after the particle swarm updates the individual extreme value and the overall extreme value process, the crossover mutation strategy is introduced; the genetic population is defined GAPOP , its population number is the same as the particle swarm population number NConsistent; define the number of population crossovers , population variation ;
[0062] Generate crossover population ;
[0063] in GA pc is the crossover population selected from the genetic population.
[0064] Generate mutant population ;
[0065] in, GA pm is the variant population selected from the genetic population, p max is the maximum value within the particle variation range, p min is the minimum value within the particle variation range, D It is the population dimension;
[0066] The final population update is: ; Complete interference suppression beamforming.
[0067] It is worth mentioning that the present invention Figure 3 As shown, Figure 3 This is a comparison of the convergence speed and accuracy of the combined particle swarm algorithm with the original algorithm (PSO) and the genetic-particle swarm algorithm (GA-PSO). The initial number of individuals in the algorithm population is 50, and the Griewank function optimization test is performed. The initial solution space distribution of CPSO is better than that of PSO and GA-PSO algorithms, and the convergence speed is the fastest, about 90% faster than PSO and about 76.7% faster than GA-PSO.
[0068] like Figure 4 As shown in the figure, the diagonal loading beamforming based on the diagonal loading factor update is compared with the traditional diagonal loading beamforming effect; in this embodiment, the simulation adopts a uniform linear array of 20 antenna elements, the signal-to-noise ratio is 10 dB, the interference-to-noise ratio is 20 dB, 3 signals are randomly generated, and the incident angle of the desired signal is selected as 0°; the interference signal is selected as -20° and 30°; under the condition of ensuring that the main beam pointing and lobe width and other performance remain unchanged, the suppression degree of UDL in the -20° direction is 4.2 dB stronger than that of the DL algorithm, and the suppression degree in the 30° direction is 4.4 dB stronger than that of the DL algorithm. It can be said that the UDL algorithm has a better interference suppression effect than the DL algorithm;
[0069] In this embodiment, the simulation uses a circular array antenna with 25 antenna elements. The number of array elements after sparseness is 12, and the sparseness rate is 44%. The layout of the sparse circular array is maintained as follows: Figure 5 As shown in Figure 1, a simulation test of the azimuth adaptive diagonal loading algorithm of a sparse circular array based on CPSO was carried out; the signal-to-noise ratio of the simulation test was 10 dB, the interference-to-noise ratio was 20 dB, three signals were randomly generated, and the incident angles of the expected signals were (180°, 45°) and (120°, 45°); the interference signal of the same channel was (240°, 45°), and the azimuth beamforming effects of the sparse circular array and the non-sparse circular array were shown in Figure 1. Figure 6 As shown in the figure, there is no significant difference in the performance between the sparse array and the non-sparse array in terms of the array main beam pointing and main beam width. In terms of interference signal suppression, the sparse array suppression performance is 10 dB (120°, 45°) and 8 dB (240°, 45°) weaker than the non-sparse array, but both can achieve good interference suppression effects. Under the premise of keeping the main beam gain unchanged, sacrificing a certain suppression depth, compared with the non-sparse array, the sparse array can still achieve good interference suppression, while reducing the design complexity and cost of the array antenna.
[0070] It must be said that in this embodiment, the number of interference signals is gradually increased to test the main beam pointing and beam width of the array, as well as the suppression effect on the interference signal; when the number of interference signals reaches three, four and five, the interference directions are (60°, 45°), (120°, 45°), (240°, 45°); (60°, 45°), (120°, 45°), (240°, 45°), (300°,45°); (30°, 45°), (60°, 45°), (120°, 45°), (240°, 45°), (300°, 45°); other conditions remain unchanged, the formation effect of the sparse circular array is as follows: Figure 7-9 shown.
[0071] In the present invention, it must be said that it is observed that as the number of interference signals gradually increases, after the sparse array is diagonally loaded, the main beam performance does not change significantly, and the suppression effect on the interference signals is gradually weakened, but interference suppression can still be achieved.
[0072] Fig.10 As shown, the azimuth beam scanning performance of the sparse circular array after using the beamforming method proposed in this article is demonstrated. It can still achieve 360° omnidirectional scanning in azimuth, and the scanning beam width remains unchanged, and it can still have the omnidirectional performance of a circular array.
[0073] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.
Claims
1. A sparse circular array interference suppression beamforming method based on a combined particle swarm optimization algorithm and diagonal loading factor update. Features: The following steps are involved: Step 1: First, according to the signal-to-noise ratio SNR, sparse circular array flow pattern , expected direction Interference signal direction , get the array receiving signal model , Expected received signal model Received signal model with interference signal ; Step 2: Calculate the noise and interference covariance matrix ; and find the search space of the formation method; Step 3: The sum of the sidelobe levels of the interference positions of the diagonally loaded beamforming is used as the evaluation function; Step 4: Use the improved combined particle swarm algorithm to optimize the initial population and design the particle swarm individual crossover mutation strategy to complete the interference suppression beamforming: The step 4 specifically includes the following steps: Step 401: Initialize algorithm parameters and set fitness function; Step 402: Generate initial population using chaotic sequence; Step 403: Enter the algorithm loop and set the speed weight update strategy; Step 404: Whether the iteration is completed, if the iteration is completed, output the population; Step 405, if the iteration is not completed, the fitness function is calculated, and the population is updated using the crossover mutation strategy to complete the interference suppression beamforming; The initial population is: In the formula a is a constant distributed over In order to improve the convergence speed of the particle swarm algorithm, a speed weight update strategy is proposed, using the weight update strategy: ; In the formula: In the formula: is called the inertia weight, is the value range of inertia weight, is the value range of the acceleration weight, is the current iteration number, is the maximum number of iterations.
2. According to the sparse circular array interference suppression beamforming method of combining particle swarm optimization and diagonal loading factor update as described in claim 1, Features: In step 2, the noise and interference covariance matrix is calculated : ; Calculated based on the interference and noise covariance matrix Take a range of values and use it as the algorithm search space: .
3. According to the sparse circular array interference suppression beamforming method of combining particle swarm optimization and diagonal loading factor update as described in claim 2, Features: In step 3, the evaluation function is: .
4. According to the sparse circular array interference suppression beamforming method of combining particle swarm optimization and diagonal loading factor update as described in claim 1, Features: In step 405: the particle swarm individual crossover mutation strategy is used to assist the algorithm to jump out of the local optimal solution.
5. According to the sparse circular array interference suppression beamforming method of combining particle swarm optimization and diagonal loading factor update as described in claim 4, Features: In step 405: after the particle swarm updates the individual extreme value and the overall extreme value, a crossover mutation strategy is introduced; a genetic population is defined GAPOP , its population number is the same as the particle swarm population number N Consistent; define the number of population crossovers , population variation ; Generate crossover population GAc : in GAP A crossover population selected from a genetic population; a mutant population is generated GAm : in, GApm is the variant population selected from the genetic population, is the maximum value within the particle variation range, yes The minimum value within the particle variation range, D is the population dimension; the final population update is: POP = [ GAc , GAm ].
