Sparsity of the Field Signal-Based Method for Improving Spatial Resolution in Antenna Sensor Array Processing

The goal of array processing is to gather information from propagating radio-wave signals, as their Direction Of Arrival (DOA). The estimation of the DOA can be carried out by extracting the information of interest from the steering vector relevant to the adopted antenna sensor array. Such task can be accomplished in a number of different ways. However, in source estimation problems, it is essential to make use of a processing algorithm which feature not only good accuracy under ideal working conditions, but also robustness against non-idealities such as noise, limitations in the amount of collectible data, correlation between the sources, and modeling errors. In this work particular attention is devoted to spectrum estimation approaches based on sparsity. Conventional algorithms based on Beamforming fail wherein the radio sources are not within Rayleigh resolution range which is a function of the number of sensors and the dimension of the array. DOA estimation techniques such as MUSIC (MUltiple Signal Classifications) allow having a larger spatial resolution compared to Beamforming-based procedures, but if the sources are very close and the Signal to Noise Ratio (SNR) level is low, the resolution turns to be low as well. A better resolution can be obtained by exploiting sparsity: if the number of sources is small, the power spectrum of the signal with respect to the location is sparse. In this way, sparsity can enhance the accuracy of the estimation. In this paper, an estimation procedure based on the sparsity of the radio signals and useful to improve the conventional MUSIC method is presented and analyzed. The sparsity level is set in order to focus the signal energy only along the actual direction of arrival. The obtained numerical results have shown an improvement of the spatial resolution as well as a reduced error in DOA estimation with respect to conventional techniques.