摘要
为了解决自适应LMS算法中收敛速度和稳态失调之间的矛盾,在选择自适应算法的步长时,通过在基于箕舌线的变步长LMS算法中引入自相关估计,对信噪比为SNR=16 dB的染噪信号进行处理.仿真结果表明:该算法使均方误差曲线在500个采样点附近达到稳态,均方误差MSE=2.595.时域波形显示,利用变步长自适应滤波算法能有效地滤除信号中的噪声,获得稳定的消噪效果.
It is a significant problem to select step-size of adaptive filtering algorithm to solve the confliction between convergence speed and steady state error. This paper presents a variable step-size LMS algorithm with autocorrelation estimation used in noise reduction. By the simulation on Matlab, the results show that MSE curve in the vicinity of 500 sampling points to achieves stable state with MSE = 2. 595. The proposed algorithm is effective to improve convergence speed and stable error, and gets the good effects of signal noise reduction.
出处
《西安工业大学学报》
CAS
2010年第1期71-74,共4页
Journal of Xi’an Technological University
关键词
变步长
自适应滤波
收敛速度
自相关估计
噪声
variable step-size
adaptive filter
convergence speed
autoeorrelation estimate
noise
