A dual parametrization approach to Nyquist filter design

In this paper, the optimum design of a factorable Nyquist filter with the intersymbol interference (ISI) being exactly zero is formulated as a nonlinear optimization problem with continuous inequality constraints. An iterative scheme is developed for solving this semi-infinite optimization problem, where an improved dual parametrization method is utilized in each iteration of the iterative scheme. Trade-off between robustness against timing jitter and small stopband attenuation is achieved via an adjustment of a parameter. Some examples are solved using the proposed iterative method.