Robust Expectation-Maximization Algorithm for DOA Estimation of Acoustic Sources in the Spherical Harmonic Domain

摘要: The direction of arrival (DOA) estimation sound sources has been a popular signal processing research topic due to its widespread applications. Using spherical microphone arrays (SMA), DOA can be applied in the harmonic (SH) domain without any spatial ambiguity. However, environment reverberation and noise degrade performance. In this paper, we propose new expectation maximization (EM) algorithm for deterministic maximum likelihood (ML) L presence spatially nonuniform SH domain. Furthermore closed-form Cramer-Rao bound (CRB) ML is derived model main idea proposed considering general received domain, reduce complexity by breaking it down into two steps: steps. reduces from 2L-dimensional space 2-dimensional space. Simulation results indicate that shows at least an improvement 6dB robustness terms root mean square error (RMSE). Moreover, RMSE very close CRB compared recent methods reverberant noisy environments large range ratio.