Low-complexity 2D DOA estimator of coherently distributed noncircular sources

In this communication, a novel low-complexity method for two dimensional (2D) DOA estimation of coherently distributed noncircular sources impinging on L-shape sensor array is proposed. Firstly, according to the noncircular property, the proposed method utilizes both the received signals and its conjugate to extend the received data matrix and form extended cross-covariance matrix. Then, we can obtain two approximate rotational invariance relation of extended steering vector by using selection matrices. At last, a novel ESPRIT-like method is employed to estimate the central azimuth angle and the central elevation angle. Numerical simulations are also provided in contrast with the other ESPRIT-like method. The results indicate that the proposed method owns better estimation performance and resolution probability in both low signal to noise ratio (SNR) and small number of sensors.