In this paper,a low complexity ESPRIT algorithm based on power method and Orthogo- nal-triangular (QR) decomposition is presented for direction finding,which does not require a priori knowledge of source number and th...In this paper,a low complexity ESPRIT algorithm based on power method and Orthogo- nal-triangular (QR) decomposition is presented for direction finding,which does not require a priori knowledge of source number and the predetermined threshold (separates the signal and noise ei- gen-values).Firstly,according to the estimation of noise subspace obtained by the power method,a novel source number detection method without eigen-decomposition is proposed based on QR de- composition.Furthermore,the eigenvectors of signal subspace can be determined according to Q matrix and then the directions of signals could be computed by the ESPRIT algorithm.To determine the source number and subspace,the computation complexity of the proposed algorithm is approximated as (2log_2 n+2.67)M^3,where n is the power of covariance matrix and M is the number of array ele- ments.Compared with the Single Vector Decomposition (SVD) based algorithm,it has a substantial computational saving with the approximation performance.The simulation results demonstrate its effectiveness and robustness.展开更多
基金Supported by the National Natural Science Foundation of China (No.60102005).
文摘In this paper,a low complexity ESPRIT algorithm based on power method and Orthogo- nal-triangular (QR) decomposition is presented for direction finding,which does not require a priori knowledge of source number and the predetermined threshold (separates the signal and noise ei- gen-values).Firstly,according to the estimation of noise subspace obtained by the power method,a novel source number detection method without eigen-decomposition is proposed based on QR de- composition.Furthermore,the eigenvectors of signal subspace can be determined according to Q matrix and then the directions of signals could be computed by the ESPRIT algorithm.To determine the source number and subspace,the computation complexity of the proposed algorithm is approximated as (2log_2 n+2.67)M^3,where n is the power of covariance matrix and M is the number of array ele- ments.Compared with the Single Vector Decomposition (SVD) based algorithm,it has a substantial computational saving with the approximation performance.The simulation results demonstrate its effectiveness and robustness.
