DESIGN OF FIR FILTER USING MODELED WINDOW FUNCTION WITH HELPING OF ARTIFICIAL NEURAL NETWORKS

Abstract

In this study the process of modeling a window function having useful spectral parameter preferred at many application area has been presented. The aim of this study is to implement designing a useful window function containing good properties of Kaiser and Ultraspherical windows widely used in literature. While Kaiser window can control mainlobe width and ripple ration characteristic but it cannot control the sidelobe roll-off ratio. The three-parameter ultraspherical window provides a much better sidelobe roll-off ratio characteristic than other windows in literature. At work, the design of a new window function that has side-lobe roll-off ratio characteristic of ultraspherical window and Kaiser window's main-lobe width and ripple ratio was obtained by helping of ANN (Artificial Neural Networks). The window obtained from the improved method was used at FIR (Finite Impulse Response) filter design. Thus it was provided to show better properties of the filter designed with window having useful spectral parameters.