Minimal sampling for direction of arrival estimation using finite rate of innovation principle

Many high resolution DOA estimation techniques have been achieved for decades, most of which depend on a large sampling. However, restricted by the sampling time or the sensor elements, the samples are usually not enough in many applications. This leads us to consider the minimal sampling for the DOA estimation, under which the DOA can not be estimated correctly. Our goal is to determine the sampling bound and to develop the corresponding estimation method. When the noise is small, we also consider the sufficient condition for the DOA estimation with different structures of the signal sources.