 |
| lojmikak |
 Wysłany: ¦ro 21:53, 26 Sty 2011 Temat postu: mbt shoes sale A steady-state error and improve th |
|
A steady-state error and improve the convergence rate of the echo cancellation algorithm
. So this is a kind of sacrifice to get a lower computational error and faster convergence of the algorithm. The most important feature of the adaptive filter is able to efficiently track time-varying unknown environment, the input signal, the output signal to reach the optimal, real-time requirements of a strong communication system, because the voice channel transmission conditions change, the system itself the beginning and can not be fully determined, but will help speed up the convergence rate of increase system stability. Adaptive filter echo cancellation system is a key technology, the convergence speed, the echo residual will be reduced. 5 Conclusion LMS algorithm because of its simplicity and ease of project implementation, has been widely used. However, slow convergence is a fixed step size LMS 3 river ice: a steady-state error and improve the convergence rate of the echo cancellation algorithm 299, the inherent shortcomings of the algorithm. This paper analyzes the LMS algorithm Two Improved Algorithms for normalized LMS algorithm LMS algorithm and the principle of delay, both algorithms reduce the error better, normalized LMS algorithm convergence rate faster than the standard LMS convergence, LMS amount of delay because of the increase in delay to better improve the output signal, reducing errors, but it will affect the amount of the delay due to the convergence speed. Combining these two algorithms, the normalized value of the control mechanism and the delay of the right combination, an improved algorithm to increase the computation as a precondition, be better able to reduce errors and ensure faster convergence speed. In real-time requirements demand high-quality and high-call systems, this improved algorithm can help reduce the noise in the IP system to reduce echo and improve the IP system,mbt shoes sale, the call quality. References [1] Ximeng He payments. ��M�� Adaptive Filter Theory. Beijing: Electronic Industry Press,ugg stivali, 2003. SimonHaykin. Adaptivefiltertheoryfourthedition [M]. Beiing: PubfishingHouseofElectronicsIndustry. 2003. [2] He Zhenya. Adaptive signal processing [M��. Beijing: Science Press,ugg italia, 2002. HEZhenya. Adaptingsignalprocessing [M1. Beijing: SciencePress. 2o02. [3�� Mawei Fu. Adaptive LMS algorithm and its application ��D��. Sichuan University, 2005. MAWeifu. TheresearchandapplicationofadaptiveLMSalgo-rithm [D]. SichunUniversity,moncler günstig, 2005. [4] DuttweilerDL. Proportionatenormalizedleastmeansquaresadaptationinechocancellers [J]. IEEETransactionsonSpeechandAudioProcessing, 2000,8 (5) :508-518. f51LoNGGZ, LINGFProakisJG. TheLMSalgorithmwithdelayedcoefficientadaptation [J��. IEEETransactionsonAcousticsSpeechandSignalProcessing, 1989,37 (9) :1397-1405. ��6] Zhang Xiang. Wang Shi. Li Jingxi. Variable step size LMS adaptive filter algorithm simulation. Computer ��J]. 2007,23 (7.1): 252.253. ZHANGXiang,Moncler jacken, WANGDeshi, LIJingxi. Thesimulationofvari-ablestep-sizeLMSadaptivealgorithm [J]. MicrocomputerInfor �� mation, 2007,23 (7-1) :252-253. [7�� DouglasSC, ZHUQ, SmithKEApipelinedLMSadaptiveFIRfilterarchitecturewithoutadaptationdelay [J]. IEEETransonSig-nalProcessing, 1998,46 (3) :775-779.
相关的主题文章：
ugg stivali Shanghai Institute of Measurement and Testing Technology (Division I
stivali ugg Canon digital camera approachable _370
moncler daunenjacke Longitudinal vibration transdu |
|
|
|
|
|