Virtual array-based spatial smoothing method for direction finding of coherent signals

This paper investigates the direction-of-arrival (DOA) estimation of coherent sources with conjugate cyclostationary properties. A virtual array-based forward backward spatial smoothing method is proposed. By properly applying multiple lag conjugate cyclic correlation functions of specified sensor outputs, a virtual array is constructed first. Based on which, the spatial smoothing technique is then introduced to estimate the DOAs of all the signals of interest. Theoretical analysis and experimental results demonstrate that, compared with the FBSS-based Conjugate Cyclic MUSIC and Conjugate Cyclic Hankel Approximation method, the proposed algorithm not only avoids the optimal lag choosing problem, but also achieves better performance in estimation accuracy. Besides, this method is robust against the difference in signal DOAs. In addition, the proposed method can also be applied to the signals with cyclostationary properties.