一種新的帶窗重疊自適應(yīng)濾波器
A Novel Overlapping Adaptive Filter with Window
-
摘要: 基于一種帶窗重疊自適應(yīng)濾波器,將重疊濾波思想引入LMS算法。利用重疊濾波的平滑性,將加窗重疊濾波和LMS算法相結(jié)合,給出了窗加權(quán)重疊LMS(WO-LMS)算法。與傳統(tǒng)的LMS算法相比,WO-LMS算法既提高了收斂速度又可以得到較低的穩(wěn)態(tài)均方誤差。理論分析了算法的收斂性,通過與LMS算法的比較,驗(yàn)證了WO-LMS算法的優(yōu)越性。
-
關(guān)鍵詞:
- 自適應(yīng)濾波器; 重疊濾波; 最小均方誤差算法; 學(xué)習(xí)曲線
Abstract: Based on a novel windowed overlapping adaptive filter, LMS algorithm with overlapping idea is introduced. In terms of the smoothness of signals processed by overlapping, a new Windowed Overlapping LMS algorithm(WO-LMS)is presented that combines the windowed overlapping idea and LMS algorithm. In comparison with traditional LMS algorithm, WO-LMS algorithm can achieve both lower steady mean square error and faster convergence speed. Convergence property is theoretically analyzed, and the experimental results prove the superiority of WO-LMS compared with LMS algorithm. -
Widrow B.Adaptive Filters. In Aspects of Network and SystemTheory, Kalman R E and Ciaris N D(Eds.)[M]. New York: Hoit,Rinehart and Winson, 1970:59 - 143.[2]Monzingo R A. Miller T W. Introduction to Adaptive Arrays.Hughes Aircraft Company Fullerton.[J].California. New York: JohnWileySons.1980,:-[3]Haykin S.Adaptive Filter Theory[M].New York: Prentice-Hall,Inc., 1986: 10- 105.[4]Jablon N K. On the Complexity of frequency-domain adaptivefiltering[J].IEEE Trans. on Signal Proceeding.1991, 39(10):2331-[5]Rafaely B, Ellion S J. A computionally efficient frequcny-domainLMS algorithm with constraints on the adaptive filter[J]. IEEETrans. on Signal Processing, 2000, 48(6): 373 - 383.[6]王兆華,韓萍,曹繼華.Fourier重疊數(shù)字濾波器[J].信號(hào)處理,2001,17(2):189-191.[7]張賢達(dá).信號(hào)處理中的線性代數(shù)[M].北京:科學(xué)出版社,1997:56-77.[8]王宏禹,邱天爽.自適應(yīng)噪聲抵消與時(shí)間延遲估計(jì)[M].大連:大連理工大學(xué)出版社,1999:66-67. -
計(jì)量
- 文章訪問數(shù): 2238
- HTML全文瀏覽量: 88
- PDF下載量: 809
- 被引次數(shù): 0
下載:
