Doppler aided blind source separation of communication signals

Blind Source Separation (BSS) algorithms have been used in previous literature to mitigate, among other problems, co-channel interference in wireless communication systems. Algorithms developed for this purpose must not only have the capability of working in the complex domain but also be able to take into account non-stationarity of sources and sensors. Analysis of commonly used source separation algorithms shows that the performance deteriorates when the sources are close to each other, which increases the error probability in blind beamforming. In this paper, by accurately modeling the source position vectors and incorporating a Doppler shift in the mixing model, we show that the separation performance of BSS algorithms considerably increases. This technique can be further extended to tracking of the sources if the Doppler shift is assumed to be constant. Demonstration of the efficacy of the algorithm on communication signals has also been studied.