PERFORMANCE ANALYSIS AND INSIGHTS INTO THE CLIPPED INPUT ADAPTIVE FILTERS APPLIED TO SYSTEM IDENTIFICATION

摘要

High computational load of conventional adaptive FIR filters applied to long system identification encourages the researchers to seek for efficient solutions for this kind of applications. One of the efficient solutions recently proposed is the threelevel clipped input LMS/RLS adaptive filter. In this paper we gain insights into such algorithms in terms of their tracking capability and their steady-state misalignment. We achieve this by analyzing the steady-state misalignment in the context of system identification for both time-invariant and timevarying systems.