A frequency domain method to estimate the coefficients of feedback control filter
Iwamatsu Yusuke, Masahiro Itou, Kensaku Fujii · The Journal of the Acoustical Society of America · 2006
This study proposes a method for repeatedly updating coefficients of a feedback control filter used for canceling a feedback path from a loudspeaker to a noise detection microphone. The coefficients are typically estimated by feeding a sequence of extra noise to the loudspeaker. However, feeding such extra noise under active noise control is undesirable. The proposed method estimates the coefficients without feeding the extra noise. Using the estimation errors involved in the coefficients of the noise control filter instead of the extra noise, the proposed method provides two independent equations and estimates coefficients of the feedback control filter by solving those. The two independent equations are provided through linear prediction using an auxiliary filter. A solution of the equations with two unknowns (a primary noise and the feedback path) provides the coefficients of the feedback control filter, perfectly canceling the feedback path. A salient drawback is that the feedback component is extremely small in comparison to the primary noise that is incident on the noise detection microphone. For that reason, estimation of the coefficients is difficult. This method solves the problem by first estimating the primary noise and then the feedback path.