Optimum finite-length space-time decision feedback equalization with multiuser detection in SDMA systems over fading channels

Abstract

We study a design of optimum finite-length space-time decision feedback equalization (STDFE) assisted multiuser detection (MUD) in space division multiple access (SDMA) systems to improve system capacity over fading channels. The MUD-STDFE receiver uses multi-dimensional finite-length anticausal symbol-spaced feedforward filter (FFF) and multi-dimensional feedback filter (FBF), where the optimality criterion is minimum mean-squared error (MMSE). This space-time DFE combined with MUD technology is capable of improving the performance significantly in a SDMA system. The algorithm for computing the tap coefficients of STDFE-MUD is presented and then applied to the fast error probability calculation using the Gauss quadrature rule (GQR), which offers accuracy as well as considerable saving of computer time.