智能优化判决反馈盲均衡算法研究

发布时间：2018-03-29 11:25

本文选题：变步长　切入点：判决反馈盲均衡　出处：《安徽理工大学》2017年硕士论文

【摘要】：盲均衡算法因其具有不需要训练序列,并且可以有效消除码间干扰(ISI)的优点,越来越多的研究学者将其作为研究的课题,提出了很多改进的盲均衡算法。本文在对常模算法(CMA)和常模判决反馈盲均衡算法(CMDFE)分析的基础上,将智能优化算法和变步长思想应用于CMDFE中,以期提高均衡效果。具体研究内容如下:1.分析了基于不同步长函数的常模盲均衡算法在对常数模盲均衡算法进行了理论分析的基础上,分析了步长对常模算法的影响。针对其收敛速度和均方误差之间的矛盾,对基于MSE的变步长常模盲均衡算法(MSE-V-CMA)和基于剩余误差自相关的变步长常模盲均衡算法(REA-V-CMA)进行了研究分析。仿真结果表明,算法在保证收敛的情况下,可以使收敛速度和均方误差都达到理想的效果。2.分析了变步长常模判决反馈盲均衡算法针对频响起伏较大的信道,分析了基于常模判决反馈盲均衡算法(CMDFE),并对算法进行了仿真。因为其仍然存在收敛速度和均方误差之间的矛盾,所以研究了基于MSE的变步长常模判决反馈盲均衡算法(MSE-V-CMDFE)和基于剩余误差自相关的变步长常模判决反馈盲均衡算法(REA-V-CMDFE),并对算法性能进行了仿真验证。3.分析了混合智能优化常模判决反馈盲均衡算法的结合针对CMDFE的初始化和局部收敛问题,在对PSO、SA、AFSA优化算法分析的基础上,研究了以下两种混合优化算法:(1)SA-PSO-CMDFE粒子群优化算法(PSO)很容易陷入局部最优,而且在收敛后期的收敛速度变慢,而模拟退火算法(SA)具有很强的全局搜索能力,可以弥补PSO在这方面的劣势,所以将SA算法与PSO算法相结合运用到CMDFE中,提出模拟退火粒子群优化常模判决反馈盲均衡算法(SA-PSO-CMDFE),通过计算机仿真验证了算法的优越性。(2)SA-AFSA-CMDFE人工鱼群优化算法(AFSA)收敛速度较慢、且较易陷于局部最优,将SA算法和AFSA算法相结合应用到CMDFE中,提出模拟退火人工鱼群优化常模判决反馈盲均衡算法(SA-AFSA-CMDFE),通过计算机仿真证明了所提出的算法具有较好的均衡性能。4.研究了混合智能优化的变步长常模判决反馈盲均衡算法为了进一步提高算法的收敛性,在前期研究工作的基础上,将变步长思想分别应用到基于模拟退火粒子群优化常模判决反馈盲均衡算法(SA-PSO-CMDFE)和基于模拟退火人工鱼群优化常模判决反馈盲均衡算法(SA-AFSA-CMDFE)中,提出基于模拟退火粒子群优化变步长常模判决反馈盲均衡算法(SA-PSO-VCMDFE)和基于模拟退火人工鱼群优化变步长常模判决反馈盲均衡算法(SA-AFSA-VCMDFE),通过仿真试验最后验证了算法有效性。
[Abstract]:Blind equalization algorithm has the advantages of no training sequence and can effectively eliminate ISI), so more and more researchers regard it as a research topic. Many improved blind equalization algorithms are proposed in this paper. Based on the analysis of the constant modulus algorithm (CMA) and the constant modulus decision feedback blind equalization algorithm (CMDFEE), the intelligent optimization algorithm and the idea of variable step size are applied to CMDFE. In order to improve the equalization effect, the specific research contents are as follows: 1. This paper analyzes the constant modulus blind equalization algorithm based on asynchronous long function, which is based on the theoretical analysis of constant modulus blind equalization algorithm. The influence of step size on the constant norm algorithm is analyzed, and the contradiction between the convergence rate and mean square error is analyzed. The variable step size constant modulus blind equalization algorithm (MSE-V-CMA) based on MSE and the variable step size constant modulus blind equalization algorithm based on residual error autocorrelation (REA-V-CMA) are studied and analyzed. The simulation results show that the algorithm can ensure convergence. The convergence speed and mean square error can reach the ideal result. 2. The variable step size constant modulus decision feedback blind equalization algorithm is analyzed for the channel with large frequency response fluctuation. This paper analyzes the blind equalization algorithm based on constant norm decision feedback and simulates the algorithm because of the contradiction between convergence speed and mean square error. Therefore, the variable step-size constant modulus decision feedback blind equalization algorithm (MSE-V-CMDFEE) based on MSE and the variable step size constant modulus decision feedback blind equalization algorithm based on residual error autocorrelation (REA-V-CMDFEE) are studied, and the performance of the algorithm is verified by simulation. The blind equalization algorithm with constant modulus decision feedback can be optimized for initialization and local convergence of CMDFE. Based on the analysis of the PSO-SA-AFSA optimization algorithm, this paper studies the following two hybrid optimization algorithms: 1 / 1, SA-PSO-CMDFE particle swarm optimization (PSOs), which are easy to fall into local optimum and slow down in the later stage of convergence. The simulated annealing algorithm (SA) has a strong global search ability, which can make up for the disadvantage of PSO in this respect. Therefore, SA algorithm and PSO algorithm are applied to CMDFE. A simulated annealing particle swarm optimization (SA-PSO) constant modulus decision feedback blind equalization algorithm (SA-PSO-CMDFEE) is proposed. The superiority of the algorithm is verified by computer simulation. The convergence speed of the algorithm is slow and the algorithm is easily trapped in the local optimum. The SA algorithm and the AFSA algorithm are applied to CMDFE. A blind equalization algorithm named SA-AFSA-CMDFEE for simulated annealing artificial fish swarm optimization is proposed. The results of computer simulation show that the proposed algorithm has good equalization performance .4.The variable step size constant modulus decision feedback blind equalization algorithm for hybrid intelligent optimization is studied. In order to further improve the convergence of the equalization algorithm, Based on the previous research work, variable step size is applied to SA-PSO-CMDFE-based particle swarm optimization and SA-AFSA-CMDFE-based simulated annealing artificial fish swarm optimization. SA-PSO-VCMDFEE algorithm based on simulated annealing particle swarm optimization is proposed, and SA-AFSA-VCMDFEE algorithm based on simulated annealing artificial fish swarm optimization is proposed. Finally, the effectiveness of the algorithm is verified by simulation experiments.
【学位授予单位】：安徽理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TN911.5

【相似文献】