Abstract
This paper investigates a new recursive least squares (RLS) algorithm for active control (AC) of impulsive noise sources. The proposed RLS algorithm is based on an objective function having robustness against the impulsive noise. This paper introduces two versions of the proposed algorithm. In the first version, the update equation comprises a fixed step-size to regulate the adaptation. Considering that a fix-valued step-size results in a trade-off situation from the view-point of the performance during the transient and at the steady-state, a second version of proposed algorithm addresses this trade-off by incorporating a convex combined step-size (CCSS) strategy. The CCSS strategy selects (automatically) a large-valued step-size at the start-up for a fast convergence speed. As the AC system approaches the steady-state, the CCSS is adapted to a small value for improving the performance at the steady-state. The simulation results show that the proposed algorithm is very effective in the various settings of AC systems for impulsive sources.