Resolving manifold ambiguities in direction finding systems

The issue of resolving manifold ambiguities in subspace-based Direction Finding (DF) systems is investigated in this paper. The cause of manifold ambiguities is due to linear dependence amongst manifold vectors. When an ambiguous situation occurs, subspace-based DF techniques fail to correctly identify a unique set of Directions-of-Arrival (DOA). This causes a performance degradation due to unreliable estimates of the parameters. In spite of the fact that there are infinitely many ambiguous scenarios in an array system, the problem in resolving manifold ambiguities has received very little attention. In this paper, two novel techniques are proposed, aiming at improving the DOA identification capability, while maintaining a minimum computational complexity. The proposed techniques adopt a beamforming class based on the Minimum Variance Distortionless Response (MVDR) criterion to estimate signal power, which is a measure for identifying the presence of a signal. Simulation results show the improved performance in terms of the identification capability over the previously proposed model fitting method.