Blind spreading sequence set estimation of the M-ary direct sequence spread spectrum signals

Abstract

In the non-cooperative context, the M-ary direct sequence spread spectrum (DSSS) signals are much more difficult to be intercepted than the conventional DSSS signals. Few literatures can be found to provide the information of this kind of signal in such circumstances. In this paper, a blind spreading sequence set estimation algorithm for the M-ary DSSS signals is proposed. This method exploits the signal structure and the cross-correlation properties between the spreading sequences. Firstly, the received signal samples are divided into a set of vectors. Secondly, using a correlation-based iteration, one of the spreading sequences is recovered. And, the vectors corresponding to this sequence are removed from the signal vector set. Thirdly, by repeating the above procedure, a set of estimates are obtained. Finally, false answers are removed by checking the estimates with the original signal samples, then, the spreading sequence set is recovered. This method is capable of providing good estimates when the signal-to-noise ratio is greater than −1dB. Besides, the proposed method is much faster and more robust than the existing K-means clustering dispreading method. Simulation results verify the capabilities of the proposed method.