Analysis of the Dereverberation Performance of Microphone Arrays

Microphone arrays and beamforming have been used extensively for speech and audio acquisition in noisy and reverberant environments. In this paper, our objective is to investigate analytically the dereverberation performance of a beamformer. As an evaluation metric we apply the commonly used direct-toreverberant ratio (DRR). Consequently, we utilize tools from statistical room acoustic theory to derive an expression for the expected DRR in the case of a delay-and-sum beamformer (DSB). Simulation results are provided to demonstrate and validate the theoretical expression.