For array processing, we consider the problem of estimating signals ofinterest, and their directions of arrival (DOA), in unknown colored noisefields. We develop an estimator that efficiently utilizes a set of noise-onlysamples and, further, can incorporate prior knowledge of the DOAs with varyingdegrees of certainty. The estimator is compared with state of the artestimators that utilize noise-only samples, and the Cram\'{e}r-Rao bound,exhibiting improved performance for smaller sample sets and in poor signalconditions.
展开▼
