Ray-Based Blind Deconvolution Combined with Optimization Algorithm

A method to improve channel impulse response (CIR) performance that combines the particle swarm optimization (PSO) algorithm with ray-based blind deconvolution (RBD) is presented. RBD is a technique that can estimate the CIRs of source to receiver array or source waveform without information about the sound source. Since the estimated CIR through RBD can be used in various ways such as the source-receiver range estimation and geoacoustic inversion, RBD have received significant attention. Unlike the conventional RBD that estimates the phase of a source waveform by applying conventional beamforming technique to the signals received in the array in the process of estimating the CIR, the method proposed in this study improves the performance of the source phase estimation by using PSO algorithm. The CIR estimation performance was improved at both short and long distances in the results of experimental data.