Localization Performance under Middle and Low Frequency Sound Source Based on Time Reversal Method in Enclosed Space

Multipath interference, reflection, sound absorption, and obstacles are important factors affect the sound source localization performance in enclosed space. Time reversal has the advantages of adaptive focusing, strong anti-reverberation capability, and the ability to overcome multipath effects, which are quite suitable for sound source localization in enclosed space due to its complex environment that many methods (e.g., beamforming, high-resolution spectral estimation, and the time delay of arrival) are not applicable. Therefore, this paper studies the localization performance under middle and low frequency sound sources cover single-frequency sinusoidal signal, bandwidth signal, and impulse signal through dual-channel matching based on time reversal method. When the frequency is below 1 kHz, the media's absorption is generally negligible. Therefore, this paper only studies frequencies within 1 kHz. In this paper, numerical simulations are used to verify the feasibility of dual-channel matching under middle and low frequency sound sources in ordinary enclosed space. Then, actual experiments are conducted in an ordinary office on the case of different sound sources. The numerical simulation and experiment results show that dual-channel matching method has satisfying localization effect. The localization effect of medium frequency signal is superior to low frequency signal. Bandwidth signals have better localization effect than single-frequency sinusoidal signals, and steady state sound source has better localization effect than impulse sound source. In general, dual-channel matching method has strong robustness and applicability for sound source localization in enclosed space.