Phase-compensator design using two-step mathematical programming

Abstract

This paper presents a new technique for designing an all-pass phase compensator (AP-PC) using a two-step mathematical programming strategy. Such an AP-PC alters the phase characteristics of a nonlinear-phase digital system and achieves phase compensation. In this paper, we first briefly review the approximate phase-error expression that is a linear function of the AP-PC coefficients, and then use the linear programming (LP) strategy to determine the sub-optimal coefficients. After getting the sub-optimal solution, we use the resulting coefficients as the initial values for the further non-linear-programming (NLP). Therefore, this design technique includes two steps (LP plus NLP: LP-NLP). Combining the above two-step optimizations significantly improves the final design accuracy. An example is provided for checking the accuracy improvement by employing the LP-NLP design strategy.