Subspace Adaptive DOA Estimation in the Presence of Coherent Signals
К. А. Сидоренко · 2013
In this paper, an adaptive implementation of the TLS-ESPRIT algorithm for angle estimation and tracking has been derived. The main ideas from adaptive subspace estimator, spatial smoothing and forward- backward averaging methods are combined to yield an estimation algorithm. A new representation of forward- backward averaging technique was offered for the usage of adaptive subspace tracking methods. Robust DOA estimation and tracking are achieved by incorporating the resulting approach with the spatial smoothing technique and TLS-ESPRIT. The simulation proves the possibility of estimating forward-backward averaged correlation matrix via proposed approach. Resulting algorithm estimates the DOA of a signal by iterations and can track the DOA in the presence of coherent signals. The simulations have shown that this algorithm can separate more than two coherent signal sources and also it represents a significant reduction in the amount of computer operations with a similar RMSE performance, when compared to a standard approach.