Multisegment optimization design of variable fractional-delay FIR filters

Abstract

This paper presents a multisegment optimization design algorithm for both even- and odd-order variable fractional-delay filters. The proposed algorithm is based on segmenting the fractional delay and separately solving the filter coefficients in each segment. Moreover, the lower bound of the polynomial order and the upper bound of the subfilter design error are derived. These bounds enable us to prove that the proposed algorithm effectively reduces the implementation complexity at the expense of storage resources. Hence, a joint search algorithm is proposed to find the lowest polynomial orders and subfilter orders that meet the filter specifications. Simulations and comparisons are conducted to illustrate the superiority of the proposed algorithm in terms of the implementation complexity over existing methods under common specifications.