A modified scheduled step-size subband active noise control algorithm with implemental considerations

Abstract

This paper proposes an algorithm that actively controls the noise using a modifi ed scheduled step-size normalized subband adaptive lter (MSSS-NSAF). One of the properties of active noise control (ANC) system is that the system has a very long FIR lter coe cients by their long echo tails. Because of the property of the ANC system, NSAF algorithm is applied to reduce the computational complexity by partitioning the input and desired signals. In adaptation step, the scheduled step-size table is applied to acheives fast convergence rate and small steady state error by the mean square deviation (MSD) analysis of the modi ed NSAF. For the calculation e ciency in chip implementation, we proposed a novel step size update algorithm. Simulation results show that the proposed MSSSNSAF algorithm has much more better convergence property than xed step-size algorithm.