DOA Estimation in Heteroscedastic Noise with sparse Bayesian Learning

摘要

We consider direction of arrival (DOA) estimationfrom long-term observations in a noisy environment. In such anenvironment the noise source might evolve, causing the stationarymodels to fail. Therefore a heteroscedastic Gaussian noise modelis introduced where the variance can vary across observationsand sensors. The source amplitudes are assumed independentzero-mean complex Gaussian distributed with unknown variances(i.e., source powers), leading to stochastic maximum likelihood(ML) DOA estimation. The DOAs are estimated from multisnapshotarray data using sparse Bayesian learning (SBL) wherethe noise is estimated across both sensors and snapshots.