The Research And Design Of The DSP-based Feedback Adaptive Noise Canceller

In the digital network age, not only can signal be received and handled speedy, but also be memoried and transmited reliably and efficiently. Digital Signal Processer of efficiency, low-cost and low-power can make those progressive to be true. With the development of the technology in the very large scale integration (VLSI) and the application of the adaptive filtering, noise cancellation studies have a theoretical and technical support. In the communication system, the use of appropriate technology of filtering to extract useful signal from the polluted signal, has a very wide application prospect.Adaptive noise cancellation technology is an expandation based on the principle of adaptive filtering. It can detect and extract the useful signal from the environment which is interfered by noise, so as to improve the quality of the useful signal. The purpose of this dissertation is to design an adaptive noise cancellation.This article had analysed the basic theory of the adaptive filter and the adaptive noise cancellation system. Index and the performance of the adaptive noise cancellation was analysed too. Two basic algorithm of the LMS and the RLS algorithm, were analysed and compared. The rationality of the system's design and and correctness of the algorithm were verified through modeling the feedback Adaptive Noise Cancellation in MATLAB and analysing the graphical of the simulation.This article had completed the design of the hardware and the software of the feedback adaptive noise cancellation system. On the fundation of the computer simulation, using the chip ADSP-BF561 of the ADI company and Intrl, this part designed a feedback adaptive noise cancellation system and the peripheral interface circuit. It also realized the LMS algorithm on the chip. Using C language it programmed and debugged in the environment of the VisualDSP++4.5. This article had provided the software flow chart of LMS algorithm and part of the core program. This design overcomes the traditional Feedforward Adaptive Noise Cancellation's shortcomings. The denoising performance of the speed and noise reduction have improved. Compared with the RLS algorithm, LMS algorithm reduced the computational complexity, and achieved good results.