This paper is based on DNCA (Dual Nested Complex Approximation) for optimizing communication channel equalizers using semi--infinite quadratic programming. The optimality criterion for the equalizer is either to minimize the complex deviation in the passband or to minimize its stopband energy when subjected to a specified peak side lobe level in the stopband. Additional linear constraints can be used to form the response by means of group delay, nulls etc. The design approach is applied to a numerical example which deals with the design of a complex communication channel equalizer.