最小均方类自适应波束形成算法的分析与比较  

Adaptive Beam forming Algorithms of Least Mean Square

作　　者：罗章凯[1] 王华力[2] 谢青[3] 孔磊[1] 王庆国[4] 

机构地区：[1]解放军理工大学通信工程学院研究生2队,江苏南京210007 [2]解放军理工大学通信工程学院 [3]解放军理工大学训练部,江苏南京210007 [4]解放军理工大学通信工程学院研究生1队

出　　处：《军事通信技术》2013年第4期33-37,共5页Journal of Military Communications Technology

摘　　要：经典最小均方(LMS)算法收敛性能与步长成正比,但仍然有很大局限性,难以实现快速收敛。针对这个问题,文章重点研究步长对算法收敛速度的影响,介绍了经典最小均方(LMS)算法、稳健变步长最小均方(RVSSLMS)算法以及最优自适应步长最小均方(OASSLMS)算法。在迭代次数相同的情况下,对三种算法的仿真进行分析比较,结果表明:三种算法都能在期望信号波达方向上形成峰值,在干扰方向上形成零陷。其中,最优自适应步长(OASSLMS)算法的权值,平方误差收敛速度最快,对期望信号的跟踪效果最好。步长优化后,权值收敛需要的迭代次数也明显减少。The convergence performance of classical Least Mean Square（LMS） algorithm is proporational to the step-size. However, LMS algorithm cannot always achieve fast convergence. To solve this problem, and with focus on the influence of step size on the convergence speed of the algorithm, three algorithms were introduced： Least Mean Square（LMS） algorithm, Robust Variable Step-Size Least Mean Square （RVSSLMS） algorithm and Optimal Adaptive Step-Size Least Mean Square （OASSLMS） algorithm. In the background of the same iteration number, simulation results were analysed and compared. The reasults indicate that all the three algorithms can form peaks in the direction of the desired signal and nulls in the direction of the interference signal. However, OASSLMS algorithm can enjoy a fast convergence speed on weights and square error, and can also achieve better performance on tracking desired signals and reduce iteration numbers.

关 键 词：最小均方 稳健变步长最小均方 自适应步长 

分 类 号：TN911.72[电子电信—通信与信息系统]