Comments on linear-phase IIR filter design using weighted LS optimization

This paper presents authors outcome of numerous experiments with linear-phase IIR filters design. The overview of direct algorithms is made and discussion on weighted least squares (WLS) method is included. Some remarks concerning efficiency of the methods are presented, in particular, we comment on the convergence ability of discussed algorithms and the influence of specification parameters on obtained filter properties. The original comparison between computational efficiency of IIR and FIR filters are also demonstrated. For some filters we evaluate cut-off value of the transition band width for which IIR filters are computationally more efficient than their FIR counterparts. This kind of information can help to choose appropriate type of filter and simplify design method.