On the performance of the variable-regularized recursive least-squares algorithms

摘要

The recursive least-squares (RLS) is a very popular adaptive algorithm, which is widely used in many system identification problems. The performance of the algorithm is controlled by two important parameters, i.e., the forgetting factor and the regularization parameter. The forgetting factor controls the "memory" of the algorithm and its value leads to a compromise between low misadjustment and fast convergence. The regularization term is required in most adaptive algorithms and its role becomes very critical in the presence of additive noise. In this paper, we present the regularized RLS algorithm and we develop a method to find its regularization parameter, which is related to the signal-to-noise ratio (SNR). Also, using a proper estimation of the SNR, we present a variable-regularized RLS (VR-RLS) algorithm.