Performance analysis of multiuser detection based receivers in fading CDMA channels

This thesis is concerned with the large-system analysis of multiuser detection based uplink direct-sequence code division multiple access (DS-CDMA) systems operating over fading channels. A number of issues are considered in this context.; When channel state information (CSI) is available at the receiver, the performance, measured in spectral efficiency and average transmitting power, is studied for systems with fixed transmission rates. When the CSI is also available at the transmitters, for systems with single or multiple transmission rates, sufficient conditions are derived for the optimality of the decoding scheme of successive interference cancellation combined with minimum mean square error multiuser detection (SIC+MMSE) in terms of minimizing the average transmitting power. When CSI is not available at the transmitters, SIC+MMSE in a group-by-group fashion is shown to be optimal in minimizing the necessary transmitting power under certain conditions. For multiple-input-multiple-output (MIMO) CDMA systems, the spectral efficiency in the low and high energy regions is obtained, which shows that adding transmit antennas does not necessarily result in better performance. Equations characterizing the tradeoff between diversity and multiplexing gains of MIMO CDMA systems are derived.; When CSI is not available at the receiver, channel estimation is necessary for coherent reception of signals. The performance of second-order-statistics based semi-blind channel estimation is studied in the large sample limit. The impact of channel estimation error on multiuser detection is analyzed via the replica method, resulting in an equivalent channel model which incorporates the imperfection of CSI. For channel-coded systems, the performance evolution of channel decoder decision feedback based iterative receivers is discussed and conditions assuring convergence to the optimal performance are derived. A low-complexity algorithm is proposed for the iterative joint mitigation of multiple access interference and intersymbol interference using decision feedback from the channel decoder, which is demonstrated to achieve good performance with reasonable computational cost.