n this paper, the application of three blind (or selfreference) spatial diversity signal processing methods to Spread Spectrum (SS) communications is described. These methods do not require any kind of side information beyond knowledge of the signal structure, in contraposition to methods that depend on training sequences. Each self-reference method is specifically designed for a particular SS transmission scheme and uses a particular signal’s structural information. All three methods, however, are derived under the common goal of finding the optimum beamformer, in the sense of maximum signal to interference-plus-noise ratio, in a blind manner. The Code Reference Beamformer for Frequency Hopping systems is an approach based on the knowledge of the hopping sequence or code structure of the signal. This method is the best alternative for array signal processing on this particular SS scheme. Two recently proposed beamforming methods, which are based on the knowledge of the redundancy structure of the desired signal, are described under the common framework of efficiently using the inherent diversity – either in frequency or time – of two signalling formats, namely, Frequency Diversity (FD) SS and DS-CDMA. The former uses frequency diversity, while the latter uses time diversity. The diversity approach presented for FDSS seems to be the best choice at the moment. Regarding DS-CDMA, for which many beamforming algorithms have been developed in the literature, the self-reference beamforming is simply an alternative blind method that shows very good performance. These diversity approaches were compared via simulations to the standard Time Reference Beamformer (TRB) and showed a performance similar to it (with 10% of the bits as a training sequence) even without the need of any side information whatsoever.

Peer Reviewed