Design of adaptive filters via sampled-data control theory

Abstract

Design and analysis of adaptive algorithms such as LMS (Least Mean Square) algorithms are conventionally done by assuming that the adaptive systems are discrete-time systems. However, in real systems such as active noise control, systems to be adapted are continuous-time ones. Therefore, such systems should be designed and analyzed as sampled-data systems. In this article, we propose a design of adaptive systems taking account of continuous-time behaviors via lifting technique in sampled-data control theory.