Personal wireless communications networks have flourished over the last decade as advances in digital cellular technology have made them more accessible to the general public. Third Generation Cellular Communication systems based on code division multiple access (CDMA) as the multiple access technique, show great scope for improvement in terms of capacity, through the use of advanced signal processing techniques. Two of the leading areas that encompass these techniques are space-time processing (smart antennas) and multiuser detection (MUD). Space-time-MUD (ST-MUD) is a relatively new field that hopes to bring together these two techniques. The focus of this thesis is ST -MUD in the context of a multi-carrier direct sequence CDMA (MC-DS-CDMA) communications system, which is one of the adopted multiple access techniques for the upcoming third generation cellular communications systems. The concepts of MUD and smart antennas are discussed, and their performance enhancing capabilities are demonstrated. The use of vector channel models and their role in modelling the propagation phenomena of the communications channel in terms of the space, time and frequency domains is also illustrated. A ST-MUD receiver architecture is presented, and the performance of the architecture with a minimum mean square error (MMSE) decision criterion is analysed in a frequency selective Rayleigh fading channel. The analysis results are verified via simulation. Three subspace MUD techniques are adapted for ST -MUD, and the joint space-frequency- multi path MMSE solution on these subspaces is given. Simulation results are used to quantify their relative performance. The relevance and applications of the subspace techniques are elaborated.

Thesis (M.Sc.Eng.)-University of Natal, Durban, 2001.