Digital Filter Design: Global Solutions via Polynomial Optimization

This paper aims at introducing a new design tool based on polynomial optimization, which can be used to obtain globally optimal designs for stable IIR filters, single-stage and multistage frequency response masking filters, and digital filters with discrete coefficients, and many other types of filters and filter banks. We show how these design problems can be formulated as polynomial optimization problems (POPs) and present a simple numerical example to demonstrate the usefulness of this new design framework