FOMP algorithm for Direction of Arrival estimation

Adaptive antennas and antenna array processing are much significant issues for improving the performance of wireless communication systems. One of the most important applications of adaptive antenna systems is the Direction of Arrival (DOA) estimation. Recently, compressive sensing algorithms, including convex relaxation methods and greedy algorithms, have been recognized as a type of novel DOA estimation methods. The Orthogonal Matching Pursuit (OMP) is an example of compressive sensing methods. Using the OMP method for DOA estimation has many advantages in comparison to other algorithms. In spite of these advantages, the DOA estimation by OMP algorithm has a substantial challenge. The OMP algorithm cannot distinguish between two adjacent signal sources. In the DOA estimation by OMP algorithm, when there are two adjacent sources, the mutual coherent condition of the compressive sensing methods is violated. The situation gets worse when there are two sources from two adjacent DOAs. In this situation, the beam former has a single peak. In this paper, we propose the Focused Orthogonal Matching Pursuit (FOMP) algorithm for estimation of DOA. The FOMP algorithm is an improved version of the OMP algorithm. It can detect two signal sources exactly when they are very close and beam former has a single peak corresponding to a direction between right directions. Simulation results demonstrate the advantages of the proposed scheme. It can be observed the FOMP algorithm could detect very close signal sources with a negligible error.