Computer-aided design of window functions for FIR filters

Linear programming is utilized to determine optimum window coefficients which depend on not only the length of the filter but also the specifications of the desired filter. With an appropriate choice of the prototype filter used to derive the initial infinite impulse response, it is demonstrated experimentally that the method can achieve design specifications with shorter length than can be obtained using the Kaiser window method. It is also shown that the performance of the proposed method is the same as that of the Parks-McClellan algorithm