A second-order cone programming approach for minimax design of 2-D FIR filters with low group delay

A design algorithm based on second-order cone programming (SOCP) for minimax design of 2-D FIR filters with low group delay is proposed. SOCP is a special class of convex programming problems that can be carried out considerably more efficiently than the popular semidefinite programming. The simulation studies presented in this paper also confirm this in a filter design context. The proposed algorithm is compared favorably with a recently proposed design method based on sequential quadratic programming. The proposed design method is expected to be a useful utility for 2-D digital filter designers whose interest is not limited to linear phase responses and low order filters.