Broadband Spatially Feedforward Active Noise Control Algorithms Using a Comb Filter

Abstract

A controller composed of a nonrecursive filter and a recursive filter is used to approximate the ideal controller for a spatially feedforward duct ANC problem. The nonrecursive part represents the dynamics of the transducer, whereas the recursive part is in the form of a comb filter. The parameters of the comb filter are obtained from the impulse response of the controller by using the least-square method. The comb filter is then cascaded with the nonrecursive part implemented as either a fixed filter or an adaptive filter. In the latter approach, two types of LMS-based algorithms are used. The proposed algorithms are implemented on the platform of a digital signal processor. Experimental results showed that the approximated controller attained 17 dB maximal attenuation in the frequency band 200∼600 Hz.