Multiple covariance matrix spectral model for direction of arrival

Abstract

Presents a novel algorithm for modeling the behavior of the measured eigenvalues in order to find the direction of arrivals (DOAs) of sources. The authors use only the eigenvalues and not the eigenvectors to find the DOAs; no other algorithm works this way. The modeling process is based on the critical distance formula developed previously by the authors, which describes the best location to add a sensor to an existing subarray in order to improve the resolution performance of an array, also it is based on the concept of a structurally adaptive array, which promotes the idea of adapting the size and geometry of a subarray (of a much larger array) in order to optimize the detection for different scenarios.