Optimization of LMS Algorithm for System Identification

TL;DR

This paper focuses on implementing and optimizing the LMS adaptive filtering algorithm for system identification, balancing performance and computational efficiency compared to RLS.

Contribution

It introduces optimization techniques for the LMS algorithm tailored for unknown system identification, emphasizing practical implementation in MATLAB.

Findings

Optimized LMS algorithm improves system identification accuracy.

Reduced computational complexity compared to RLS.

Effective in real-world adaptive filtering scenarios.

Abstract

An adaptive filter is defined as a digital filter that has the capability of self adjusting its transfer function under the control of some optimizing algorithms. Most common optimizing algorithms are Least Mean Square (LMS) and Recursive Least Square (RLS). Although RLS algorithm perform superior to LMS algorithm, it has very high computational complexity so not useful in most of the practical scenario. So most feasible choice of the adaptive filtering algorithm is the LMS algorithm including its various variants. The LMS algorithm uses transversal FIR filter as underlying digital filter. This paper is based on implementation and optimization of LMS algorithm for the application of unknown system identification. Keywords- Adaptive Filtering, LMS Algorithm, Optimization, System Identification, MATLAB