Blind Source Separation, ISI Cancellation and Carrier Phase Recovery in SDMA Systems for Mobile Communications

Current rates of demand for radio cellular services will soon reach the saturation point of the actual cellular infrastructures. Thus, spectral bandwidth-saving multiple access schemes are necessary to expand the actual networks' capacity. In the Spatial Division Multiple Access (SDMA) approach, users within the same cell share the same time/frequency channel and the base station receiver exploits the spatial dimension of the radio resource to remove crosstalk. In this paper, we present a blind multi-user equalization technique for SDMA architectures which performs the separation of digital sources, in the presence of intersymbol interference (ISI) and phase carrier drifts. We take advantage of the sources' joint finite alphabet and statistical properties, and use only second-order statistics in order to produce valid channel estimates based on short data packets. Computer simulations using the TU-GSM multipath channel model for radio communications in urban environments demonstrate the potential of our technique.