Quasi-Newton simultaneous perturbation stochastic approximation algorithm for broadband active noise control (L)

This Letter proposes an active noise control (ANC) algorithm that integrates the simultaneous perturbation stochastic approximation (SPSA) algorithm with the quasi-Newton method to enhance broadband noise reduction. The proposed algorithm efficiently estimates the gradient of the cost function and simultaneously updates both the control filter and the inverse Hessian matrix. Its effectiveness is validated using in-vehicle noise data. Simulation results demonstrate that the proposed ANC algorithm achieves faster convergence and a lower steady-state mean square error compared to existing SPSA methods, with only a marginal increase in computational complexity.