Recursive Least Complex Signa Signum Algorithm

摘要

In this paper, we first review Complex Signo Signum Algorithm (CSSA, and then combine the CSSA with a recursive least estimate of the inverse covariance matrix of the regressor, proposing Recursive Least Complex Signa Signum Algorithm (RLCSSA. Two types of impulse noise are considered: one is found in observation noise at the filter output and another at the filter input. Statistical analysis of the RLCSSA is developed for calculating theoretical convergence behavior. Through experiments with typical examples, we demonstrate effectiveness of the proposed algorithm in accelerating the convergence speed, while preserving the robustness of the CSSA against both types of impulse noise. The calculated theoretical convergence curves are generally in good agreement with the simulated ones that shows the validity and accuracy of the analysis.