The Study Of Analysis And Methods Of Adaptive Noise Cancellation In Communication Systems

Signal transmission will be exposed to many kinds of noise or interference pollution in communication systems, it is always an urgent problem that how to extract useful signal from being contaminated effectively in communicational technology. The usual approach is to allow the contaminated signal pass through a filter; the purpose is to restrain noise or interference while retaining the useful signal. This filter is divided into classic and modern filter, the characteristic of the classic is useful input signal frequency components and want to filter out frequency components in different frequency bands, we can achieve the purpose to through an appropriate selected frequency filter; the characteristic of modern filter is able to signal interference and the frequency bands overlap each other under the effective filtering, such as Wiener filter, Kalman filter, adaptive filter, etc..This paper has made a brief introduction of the noise in communicational systems, adaptive filtering theory and signal processing technology; analyzing and studying the basic principles of adaptive filter and adaptive noise cancellation system, the technical indicators of measuring adaptive noise cancellation system, the infection of noise correlation on adaptive noise cancellation system and the performance of adaptive filter on adaptive noise cancellation system. Focused on LMS algorithm, analysis the advantages and disadvantages of many algorithms, such as the fixed step size LMS algorithm, normalized LMS algorithm, the join of the momentum LMS algorithm, Sigmoid function which based on the LMS algorithm. On this basis, a variable step size based on the cos(θ)LMS algorithm is proposed, and modify the original algorithm. Make a simulinking research of these algorithms on the platform of MATLAB , mainly including, (1) the noise reduction capacity of algorithm; (2) convergence rate of algorithm, steady-state error and other performance indicators; (3) the selection of algorithm's parameters. Confirming the given conclusions by the simulation of the theoretical analysis.