Deterministic Algorithms for Four-Dimensional Imaging in Colocated MIMO OFDM-Based Radar Systems

In this manuscript, the problem of detecting multiple targets and jointly estimating their spatial coordinates (namely, the range, the Doppler and the direction of arrival of their electromagnetic echoes) in a colocated multiple-input multiple-output radar system employing orthogonal frequency division multiplexing is investigated. It is well known its optimal solution, namely the joint maximum likelihood estimator of an unknown number of targets, is unfeasible because of its huge computational complexity. Moreover, until now, sub-optimal solutions have not been proposed in the technical literature. In this manuscript a novel approach to the development of reduced complexity solutions is illustrated. It is based on the idea of separating angle estimation from range-Doppler estimation, and of exploiting known algorithms for solving these two sub-problems. A detailed analysis of the accuracy and complexity of various detection and estimation methods based on this approach is provided. Our numerical results evidence that one of these methods is able to approach optimal performance in the maximum likelihood sense with a limited computational effort in different scenarios.