On a Multiplierless FIR Decimation Filter Design

This paper presents one method for a multiplierless FIR decimation filter design. The method is based on the IFIR structure and the rounding and sharpening techniques. It is supposed that the decimation factor can be presented as the product of two factors. In that way the expanded model filter can be moved to the lower rate. The coefficients of the model and interpolator filters are rounded to the nearest integers and the sharpening technique is applied to the model filter in order to improve the magnitude characteristics of the overall filter. The resulting filter is a multiplierless filter because the integer multiplications can be realized by only additions and shifts. Additionally the polyphase decomposition is applied to the rounded interpolator and its polyphase components are moved to the lower rate. Consequently, there is no filtering at high input rate.