Real-time sound source localisation for target tracking applications using an asynchronous microphone array

This paper presents a strategy for sound source localisation using an asynchronous microphone array. The proposed method is suitable for target tracking applications, in which the sound source with a known frequency is attached to the target. Conventional microphone array technologies require a multi-channel A/D converter for inter-microphone synchronization making the technology relatively expensive. In this work, the requirement of synchronization between channels is relaxed by adding an external reference audio signal. The only assumption is that the frequencies of the reference signal and the sound source attached to the target are fixed and known beforehand. By exploiting the information provided by the known reference signal, the Direction Of Arrival (DOA) of target sound source can be calculated in real-time. The key idea of the algorithm is to use the reference source to “pseudo-align” the audio signals from different channels. Once the channels are “pseudo-aligned”, a dedicated DOA estimation method based on Time Difference Of Arrival (TDOA) can be employed to find the relative bearing information between the target sound source and microphone array. Due to the narrow band of frequency of target sound source, the proposed approach is proven to be robust to low signals-to-noise ratios. Comprehensive simulations and experimental results are presented to show the validity of the algorithm.