DOA estimation of multiple moving sources using recursive EM algorithms

The paper deals with maximum likelihood (ML) direction of arrival (DOA) estimation of multiple moving sources. The time-varying DOAs are described by a linear polynomial. Based on the recursive EM algorithm, we develop a recursive procedure to estimate the coefficients of the corresponding polynomial. Because of the simple structure of the augmented data, the resulting recursion has a very simple implementation. Compared to the recursive algorithm developed by L. Frenkel and M. Feder (see IEEE Trans. Sig. Processing, vol.47, no.2, p.306-20, 1999), our approach has the advantage of being insensitive to signal variations. In addition, extension to the wideband case is straightforward. Simulation results show that closely spaced signals can be successfully tracked by the proposed method.