The Research And Application Of Adaptive Filtering Algorithm

Adaptive filtering algorithm, which decides directly the performance of filtering, is seemed as the important part of the adaptive fiter. Presently the research on it is one of the most active tasks. The goal of the reaearchers'endeavor is to find an adaptive filtering algorithm, which owns fast convergence speed, low calculated amount and good robustness. Based on the basic adaptive filtering principle, firstly, this paper introduces the present main adaptive filtering algorithms and their applications. Especially the LMS algorithm and NLMS algorithm are deeply analyzed. Secondly, this paper introduces several typical variable step size LMS and NLMS algorithms, and compares and evaluates their performance. Finally, the paper presents several kinds of modified variable step size LMS and NLMS algorithms. The computer simulations under the condition of system identity and adaptive channel equalization confirm the better performance of these algorithms, compared to other algorithms.