Robust Active Noise Control: Minimum Output Variance Approach With Least Mean Lp-Norm Algorithm

The filtered-x least mean <inline-formula> <tex-math notation="LaTeX">${\mathcal {([A-Z]+)}}_{([a-z]+)}$ </tex-math></inline-formula>-norm (FxLMP) algorithm, renowned for its widespread usage, exhibits superior performance over the filtered-x Least Mean Square (FxLMS) algorithm in impulsive noise-afflicted environments. Nevertheless, the FxLMP algorithm experiences performance degradation when confronted with nonlinearity in either the primary or secondary path. In response to this challenge, we propose a novel algorithm known as the minimum output variance filtered-s LMP (MOV-FsLMP), which effectively tackles nonlinear active noise control problems by imposing constraints on variance or power. Notably, this proposed algorithm exhibits improved convergence and stability even in impulsive noise environments, all while maintaining lower computational complexity. Extensive simulations validate the efficacy of the proposed MOV-FsLMP method, surpassing that of state-of-the-art alternatives.