摘要
针对已有最小均方(LMS)自适应干扰对消算法初期收敛速度慢且整体误差偏大的问题,提出一种改进变步长LMS算法。该算法在保证收敛速度和误差精度要求下,简化了步长因子μ(n),且采取分段迭代的方法,克服算法初期收敛慢的缺点。在误差|e(n)|较大时,文中提出通过调整系数ε调节步长μ(n)的变化速度,增强了迭代稳定性。仿真结果表明:与正态分布曲线步长等算法相比,使用文中算法后,宽带与窄带电磁干扰信号的收敛速度至少提高13%,干扰对消后期望信号的输出信噪比至少提高26%。该算法可用于复杂电磁环境下的电磁干扰消除。
Aiming at the problem of slow convergence speed and large overall error at the beginning for the existing least mean square(LMS)adaptive interference cancellation algorithm,a modified variable-step LMS algorithm is proposed.The algorithm simplifies the step size factorμ(n)under the requirement of convergence speed and error accuracy,and adopts the segmented iteration method to overcome the disadvantage of slow convergence at the beginning of the algorithm.When the error|e(n)|is large,the step sizeμ(n)change rate is adjusted by the coefficientε,which enhances the iterative stability.Simulation results show that the convergence speed of the algorithm in this paper for broadband and narrowband electromagnetic interference signals is improved by at least 13%compared with algorithms such as the normal distribution curve step,and the output signal-to-noise ratio of the desired signal after interference cancellation is improved by at least 26%.The algorithm in this paper can be used for electromagnetic interference cancellation in complex electromagnetic environments.
作者
陈东伟
李明洁
方庆园
刘卫东
CHEN Dongwei;LI Mingjie;FANG Qingyuan;LIU Weidong(Hebei Provincial Key Laboratory of Electromagnetic Environmental Effects and Information Processing,Shijiazhuang Tiedao University,Shijiazhuang 050043,China)
出处
《微波学报》
CSCD
北大核心
2024年第4期30-35,共6页
Journal of Microwaves
基金
国家自然科学基金(61801309)。
关键词
LMS算法
改进变步长
干扰对消
调整系数
分段迭代
LMS algorithm
improved variable step size
interference cancellation
adjustment factor
segmented iteration
