Minimax design of 2-D IIR digital filters using sequential semidefinite programming

A method for the minimax design of quadrantally symmetric 2-D IIR filters with guaranteed stability is proposed. The design problem is solved by formulating an objective function in the semidefinite programming framework using a linear approximation for the transfer function. The issue of filter stability is addressed by converting the stability constraints into linear matrix inequalities based on the well-know Lyapunov stability theory.