Robust equalizer design for adaptive room impulse response compensation

Abstract

Adaptive room equalization aims at providing a listener with an audio experience, which is very close to the original audio signal. The equalizer, which is an adaptive filter, compensates for the disturbance in the audio signal contributed by the impulse response of the room. One of the most popular algorithms employed for the design of an adaptive room equalizer is the filtered-x improved proportionate normalized least mean square (Fx-IPNLMS) algorithm. IPNLMS is effective in equalizing sparse as well as non-sparse room impulse responses. However, Fx-IPNLMS is not robust to strong disturbances picked up by the microphone used in the equalization process and the algorithm may even diverge in such scenarios. With an objective to overcome this limitation of IPNLMS based adaptive room equalization schemes, a robust Fx-IPNLMS algorithm has been developed in this paper. The new algorithm has been shown to provide robust room equalization and thus an enhanced audio experience.