DOA estimation in conformal arrays based on the nested array principles

The nested array structure has attracted great attention recently due to its ability in reducing the number of sensors in an array and at the same time preserving the array performance. While a uniform linear array (ULA) can detect at most N - 1 sources with N sensors, a nested array can provide O(N-2) degrees of freedom with the same number of sensors; allowing us to detect K sources with K > N sensors. Direction of arrival (DOA) estimation in a conformal array is a challenging task. In this article, by breaking the conformal array into smaller sub-arrays and using an interpolation technique, we employ the nested array principles to detect more number of sources than sensors. This comes at the cost of more snapshots and lower resolution, in the DOA estimation of an arbitrarily-shaped conformal array. Each sub-array in the conformal array is selected such that the "shadow effect" which leads to an incomplete steering vector in the DOA estimation algorithm is eliminated. The selected sub-arrays are then transformed to virtual nested arrays where more degrees of freedom can be obtained by applying the MUSIC algorithm for DOA estimation. The application of our proposed method is highlighted by considering a set of comprehensive examples for cylindrical and spherical arrays. (C) 2015 Elsevier Inc. All rights reserved.