DOA estimation after blind identification of subspace channel vectors in multi-path contexts

Over the two last decades. a large number of narrow band direction finding techniques have been developed in order to estimate the DOAs (Directions Of Arrival) of impinging sources received on an array of sensors. To improve the classical methods in ambiguity robustness and accuracy. the blind-MUSIC method [51 has been proposed. It applies a separated DOA estimation of each independent source after a blind steering vector identification. In a radiocommunication context. the Signal of each emitter propagates through different paths and thus creates correlated sources at the receiver. The purpose of this paper is to apply a separated DOA estimation of independent emitters in a multi-path context taking into account both the large number of sources and the correlation due to the multi-path propagation. The approach is based on 2 steps: firstly a blind identification [2] [3] of emitters subspace channel is made and secondly a DOA estimation is applied to each estimated subspace channel. Using this new method. more sources than sensors can be identified.