A Novel Self-Calibration Algorithm for Coherent Signal Sources

To estimate direction of arrival (DOA) for coherent signal sources in the presence of mutual coupling, a novel self-calibration and decorrelation method was presented based on the first order statistics of array received data. A pseudo covariance matrix was constructed by the first order statistics. Theoretical analysis proves that, the rank of the matrix was determined by the number of signals, without any relation with the coherency of signals. The mutual coupling coefficients have already been isolated from the ideal steering vectors. This shows that DOA estimation can be achieved by performing once decomposition to the reconstructed matrix. Moreover, the performance of subspace estimation and the fuzziness of DOA estimation caused by mutual coupling coefficients were analyzed respectively. Results show that, the proposed method has high performance under scenarios of low SNR and snapshot deficiency with simple implementation and low computational complexity. Simulations present the effectiveness of the method.