Performance Analysis of a Simplified RLS Algorithm for the Estimation of Sinusoidal Signals in Additive Noise

摘要

Adaptive estimation of nonstationary sinusoidal signals or quasi-periodic signals in additive noise is of essential importance in many diverse engineering fields, such as communi- cations, biomedical engineering, power systems, pitch detection in transcription and so forth. So far, Kalman filtering based techniques, recursive least square (RLS), simplified RLS (SRLS) and LMS algorithms, for examples, have been developed for this purpose. This work presents in detail a performance analysis for the SRLS algorithm proposed recently in the literature, which is used to estimate an enhanced sinusoid.