Improved Toeplitz Algorithms to Coherent Sources DOA Estimation

In this paper, an improved Toeplitz matrix construction algorithm of covariance matrix is applied. This covariance matrix can be used to estimate both the incoherent and the coherent source direction with the Multiple Signal Classification (¿MUSIC¿) algorithm. First, the proposed algorithm uses the correlation sequences of output narrowband signals between each sensor and the reference sensor instead of the corresponding output sequence; Second, the inner product operation is carried on to the new sequences of each sensor and the reference sensor; Last, we construct a Toeplitz matrix using the inner product sequence. Obviously, the constructed covariance matrix for ¿MUSIC¿ algorithm is a non-singular matrix no matter the source is coherent or not. Furthermore, a comparative Monte Carlo simulation test among MUSIC, TOP and this paper's improved Correlation Toeplitz (CTOP) Method is presented on linear equip spaced array (LEA) carrying on to both the incoherent and the coherent source signals under each kind of signal-to-noise ratio (SNR). All results show that the CTOP has more accurate estimation performance than MUSIC and TOP method to both the incoherent and the coherent source, the CTOP has superb performance under the low SNR condition, and what is more commendable is that the CTOP is an unbiased estimation method and can overcome the influence of sensor's error also.