Study And Design Of Adaptive Noise Canceller Based On DSP

The adaptive noise canceller is a kind of application based on adaptive filtering technique, it can eliminate the noise interference effectively in a time-varying and non-stationary environment, and has been widely used in adaptive signal processing. With the development of large scale digital integrated circuit and computer technology, especially the appearance of high performance digital signal processor, has provided hardware support for adaptive noise canceller. Based on the study of adaptive filtering technique and related algorithms, this paper is committed to building an adaptive noise cancellation hardware platform based on DSP, and applying the platform to speech signal denoising in actual noise environment.This article introduces the method which combines the theoretical analysis, software simulation with the hardware design. Based on the adaptive algorithm, the software simulation results verify the influence of correlation parameters to the adaptive algorithm performance. Then apply the adaptive algorithm to the adaptive de-noising experiment through modeling simulation and compare the performance of the algorithm based on the signal de-noising result. Finally, accomplish the hardware design of adaptive noise canceller through the adaptive noise cancellation hardware platform, which is established based on DSP. The main work of this article is as follows:1. Study on the adaptive algorithm, focus on the derivation process of the algorithm of LMS and RLS, and analyze the factors influencing the performance of LMS algorithm. Because of the contradictory requirements to the step size factor caused by the convergence speed and stability maladjustment of LMS, we study on several variable step size algorithm, and propose an improved algorithm based on this.2. The improved variable step size LMS algorithm is applied to adaptive noise cancellation system simulated on MATLAB, the result show that it has remarkable feature and can improve the performance of the system.3. With the SIMULINK platform, build up the real time adaptive noise cancellation systems based on LMS algorithm and RLS algorithm respectively, and confirm the optimal parameters of the algorithms through the analysis of the simulation results.4. Based on the DSP chip produced by the TI company, this article construct a minimum hardware system of adaptive noise canceller, and accomplish the hardware design and debugging. Also a speech signal de-noising test was designed to verify the function of the system.