Abstract:
State-space recursive least-squares (SSRLS) allow the designer to choose an appropriate model. This attribute of SSRLS suits the model dependent nature of the tracking pr...Show MoreMetadata
Abstract:
State-space recursive least-squares (SSRLS) allow the designer to choose an appropriate model. This attribute of SSRLS suits the model dependent nature of the tracking problem. On the other hand, the standard RLS and LMS assume a multiple linear regression model. Therefore, their tracking abilities are limited. In this paper, we begin with the derivation of time-varying SSRLS which is followed by some related details. Our major contribution is the development of versatile algorithms that can efficiently track time-varying SSRLS which is followed by some related details. Our major contribution is the development of versatile algorithms that can efficiently track time-varying systems. Superior tracking performance of SSRLS is demonstrated by a couple of examples in the end. The paper provides a guideline that would enable a designer to devise newer algorithms for a wide range of problems.
Date of Conference: 23-26 May 2004
Date Added to IEEE Xplore: 03 September 2004
Print ISBN:0-7803-8251-X