A Solution to the Narrow-Band Coherency Problem in Multiple Source Location

In this correspondence, we propose a new approach for estimating the directions of arrival (DOA's) of narrow-hand coherent signals. The proposed method is based on constructing the virtually translated steering matrix, using the basic concept originally proposed for wide-hand signals by Wang and Kaveh. Simulation results are il- lustrated to examine the comparative performance of this approach relative to the spatial smoothing technique and the Cramer-Ran lower bound (CRLB) and demonstrate the effectiveness of the proposed method in a general array with the various choice of the array dis- placement vectors. The performance of superresolution algorithms is severely de- graded when some of the incident signals are perfectly correlated (coherent) or just highly correlated, since the source covariance matrix is singular or almost singular. Herein, we propose a new approach which can estimate the di- rections of arrival (DOA's) of narrow-band coherent signals inci- dent on a general array using the concept of signal subspace trans- formation in the spatial domain. The basic idea of the proposed method is similar to that of Wang and Kaveh's CSM technique ( 11 in terms of approximating the transformation matrix by using initial angle estimates. However, while the CSM technique in wide-band processing combines signal subspaces at different frequencies into one coherent signal subspace, our proposed method transforms the steering matrix associated with the original array to those of the virtually translated arrays. Furthermore, contrary to Wang and Ka- veh's CSM technique, the efficiency of signal decorrelation in our algorithm is not affected by phase delays between coherent signals. In Section 11, a signal model is formulated for coherent problem. Section 111 presents the signal decorrelation via virtual translation of array (SDVTA) algorithm for estimating the angles of multiple narrow-band coherent signals. Simulation results are illustrated in Section IV in order to demonstrate the effectiveness of the proposed algorithm for resolving narrow-band coherent signals.