Adaptive subspace algorithm for direction finding and tracking

This paper presents an adaptive subspace-based method for the direction of arrival estimation and the tracking of multiple sources. This method relies on a linear operator, referred to as the Propagator, which only exploits the linear independency of the source steering vectors, and which allows the determination of the source and the noise subspaces without any eigendecomposition ofthe cross-spectral matrix of the received signals. Two gradient-based adaptive algorithms are here proposed for the estimation of the Propagator, and then of the source and the noise subspaces. A theoretical analysis of the behavior of these two algorithms in a nonstationary environment is given. Simulations exhibit the good performances of the proposed algorithms for tracking moving sources.