Normalized Kernel Risk-Sensitive Loss Algorithm for Active Noise Control

In recent years, active noise control (ANC) has attracted considerable research interests, and a number of adaptive algorithms have been developed for enhancing the performance of ANC systems.However, it should be noted that some algorithms exhibit unsatisfactory steady-state performance when confronted with non-Gaussian noise, particularly impulsive noise.To address the above issue, the filtered-x kernel risk-sensitive loss (FxKRSL) algorithm is proposed, which adopts KRSL as the optimization criterion.In this paper, in order to further enhance the robustness of FxKRSL, we develop the filtered-x normalized kernel risk-sensitive loss (FxNKRSL) algorithm, which incorporates the normalized method.Simulation results confirm the superior performance of the FxNKRSL algorithm for noise control in non-Gaussian noise environments. CCS Concepts Theory of computation Mathematical optimization; Design and analysis of algorithms; Computing methodologies Modeling and simulation.