| This thesis describes the design and implementation of the adaptive filters on Texas Instruments (TI) TMS320C31 Digital Signal Processor (DSP). A Correlation least mean square (LMS) algorithm is used in the design of the adaptive filter. The application of adaptive filter in noise cancellation is to produce an optimized estimate of the noise and signal. Adaptive filter is necessary when a specific signal output is desired but the coefficients of that filter cannot be determined at the output. This may be due to genuine uncertainty about the incoming signal or because it changes slightly with time or environment. An adaptive filter is one, which has coefficients, which are designed to vary with time in a tightly controlled manner, and intervention of the user. Assembly code is written based on the TMS320C31 architecture and is loaded into DSK (DSP Starter Kit) and the performance of the filter is analyzed for different input noises by changing adaptation rate, smoothing factor and scale factor. |
