Direction of Arrival Estimation of Digital Sources With Uni-Vector-Sensor ESPRIT

This paper addresses the multi-source parametric estimation problem involving a single electromagnetic vector sensor. The Uni-Vector-Sensor ESPRIT algorithm was originally devised for the harmonic signal model to estimate the direction of arrival (DOA) and polarization of up to five uncorrelated sources. Here we show that this algorithm may be applied to the case of digitally modulated source signals. The performance with this signal model is comparable to the case of harmonic sources. However, this requires that the time lag between data sets in the Uni-Vector-Sensor ESPRIT algorithm to be small relative to the symbol periods of the digital signals. This is necessary for high correlation of the signals between the data sets, but the delay must still be large enough to maintain distinct phase offsets between the sources. The performance of the DOA estimation algorithm for digitally modulated sources and its limitations were verified through simulations and compared to other algorithms.