Design of new filterbanks by using window functions

This thesis study presents a new method for the design of M-channel cosine modulated filter banks. In this proposed method, the exponential window function having two independent parameters is used for the design of prototype FIR (Finite Impulse Response) filter of the filter banks. To minimize the amplitude and aliasing errors that occur in the design of the filter banks, a fast and simple cost function optimizing the cut-off frequency is used. The performance of the proposed system has been analyzed by making designs for different channel numbers, minimum stopband attenuations, and filter lengths. As a design example, a cosine modulated filter bank having 32-channel and a stopband attenuation with 100 dB is designed and analyzed. In addition, in order to test the performance of the method comparatively, the filterbank designed by the proposed method is compared with the one designed by the Kaiser window based Lin's method. The simulation results show that the filter bank designed by the proposed method has better design performance in terms of the amplitude and aliasing errors. Moreover, the proposed method is compared in terms of the amplitude and aliasing errors with the Kaiser window based Lin's method for different channel numbers, minimum stopband attenuations, and filter lengths and by applying both methods for ECG signals their subband processing performances are investigated.