Adaptive BLAST-type decision-feedback equalizers for DS-CDMA systems

In this paper, we propose two new adaptive equalization algorithms for direct sequence code division multiple access (DS-CDMA) systems operating over time-varying and frequency selective channels. The equalization schemes consist of a number of serially connected stages and detect users in an ordered manner, applying a decision feedback equalizer (DFE) at each stage. Both the equalizer filters and the order in which the users are extracted are updated in a recursive least squares (RLS) manner, efficiently realized through time- and order-update recursions. V-BLAST detection ordering is implemented, that is, the stronger signal is extracted first so that the weaker users can be more easily detected. The spreading codes are unavailable at the receiver of the first scheme, whereas the second algorithm employs the RAKE receiver concept, incorporating knowledge of the spreading sequences to offer performance improvement. The bit error rate (BER) performance of the equalizers is evaluated via simulations, in both mild and severe near–far environments. Their superiority over existing techniques is demonstrated.

► Adaptive equalization for DS-CDMA systems is considered.
► Adaptive BLAST-type DFE's are adapted to this context, through a MIMO formulation of the DS-CDMA system.
► Two schemes are presented, based on efficient RLS update of both the DFE filters and the optimal detection order.
► The algorithms exhibit significant performance gains over existing approaches.
► The effects of several parameters on the performance are assessed.