Correlation Maximization-Based Sampling Rate Offset Estimation for Distributed Microphone Arrays

In this paper, we investigate the sampling rate mismatch problem in distributed microphone arrays and propose a correlation maximization algorithm to blindly estimate the sampling rate offset between two asynchronously sampled microphone signals. We approximate the sampling rate offset with a linear-phase drift model in the short-time Fourier transform (STFT) domain and show that the correlation coefficient between two microphone signals tends to present the highest value when the sampling of the two microphone signals is synchronized. Based on this finding we propose the correlation maximization algorithm, which performs sampling rate compensation on two microphone signals with different possible offset values and calculates their correlation coefficient after compensation. The offset value that leads to the largest correlation coefficient is chosen as the optimal estimate. Since the precision of the STFT linear-phase drift model used in the algorithm degrades as the sampling rate offset or the signal length is increased, we further propose a two-stage exhaustive search scheme to detect the optimal sampling rate offset. This scheme is able to minimize the influence of the linear-phase drift model error in order to improve the sampling rate offset estimation accuracy. Both simulated as well as real-world experiments confirm the effectiveness of the proposed algorithm.