An Avoidanvce of Local Minimum Stagnation in Designing IIR Filters Using SOA

Abstract

In this paper, we propose a design method of IIR(Infinite Impulse Response) filters using SOA(Seeker Optimization Algorithm). The design problem of IIR filter is generally a complex approximation problem in which both a magnitude response and a phase response are approximated simultaneously. In addition, a stability of the filters must be considered. Therefore, the problem falls into a nonlinear optimization problem. Recently, the SOA has been proposed as one of the heuristic methods which can solve the nonlinear optimization problems approximately. The SOA is based on the swarm optimization, and inspired by a human activity in seaching anything. In our method, the SOA is applied to the IIR filter design and a step size used in the SOA is proposed to avoid for stagnating around local minimums. Several design examples are shown to present the effectiveness of the proposed method.