Minimax design of IIR digital filters using partial second-order factor updates

In this paper, a novel method is developed for the design of IIR digital filters in the minimax sense. The Steiglitz-McBride (SM) scheme is adopted in the proposed method to simplify the design procedure. But, different from traditional SM-based design approaches, our method decomposes the denominator polynomial of an IIR digital filter as a cascade of several second-order factors (SOFs) and a higher-order factor (HOF). The SOFs are essentially detached from the HOF so as to control the positions of poles close to the boundary of a specified stability domain. This design model can significantly enhance the performance of the design strategy that considers the denominator polynomial as a whole and simultaneously updates all the poles of an IIR digital filter. On the other hand, the proposed method can greatly improve the computational efficiency of SM-based design methods which separately update each SOF. Another merit of the proposed method is that it can automatically determine the number of SOFs, which makes the proposed method competent to various design specifications. Simulation results demonstrate the effectiveness of the proposed method and its advantages compared to other traditional SM-basSM-baseded design methods.