An Improved Variable Step Size LMS Adaptive Filtering Algorithm

An new improved variable step size LMS algorithm based on analyzing the traditional LMS algorithm and its improved algorithm are presented.The error signal and the error signal correlation value are used to adjust step,in order to overcome the general variable step size LMS algorithm's shortcomings of poor anti-noise performance under low SNR and slow convergence speed under high SNR.The computer simulation results show that,as compared with traditional LMS algorithm and VSSLMS algorithm,this algorithm converges faster,with smaller mean square error and good anti-noise performance.