Blind multi-user detectors exploiting spatial diversity in DS-CDMA downlink

We propose second-order-statistics (SOS) based spatial diversity exploiting the subspace methods to blindly identify the downlink channel of direct sequence-code division multiple access (DS-CDMA) systems. Once the channel has been identified, a linear receiver is used to recover the transmitted symbols. The new techniques supports 100% loading. It is shown that base on the identified signal subspace and the channel parameters, linear detectors can be constructed in closed form at a reduced computational cost.