Intelligent receivers for multi-user CDMA systems

In DS-CDMA system, overcoming the problem of multiple access interference (MAI) and multi-path fading is imperative to improve the performance. A number of multi-user detection (MUD) algorithms have been proposed to mitigate the MAI. This report simulates the performance of two MUD schemes known as successive interference cancellation (SIC) and parallel interference cancellation (PIC) and compares the performance with some linear MUD algorithms such as correlating, decorrelator, minimum mean-squared error (MMSE) and maximum likelihood detection (MLD). The simulation results illustrate that SIC and PIC can achieve better performance but reduce the computation complexity. Channel estimation and spatial diversity techniques are investigated in this report to combat the multi-path fading. A reduced combined multiuser channel estimation and detection algorithm has been explored, which shows to have more efficient implementation meanwhile maintain the required performance. Spatial diversity technique such as multiple transmit and/or receive antennas can be used to mitigate the effects of the multi-path fading but it need several antennas. Multi-relay system introduced in this report enables single antenna mobile to benefit from the MIMO system and has been proofed to be able to achieve full diversity order. From our study we can draw the conclusion that both these SIC and PIC algorithms, maximum likelihoodbased channel estimation scheme and the spatial diversity techniques are quite effective in improving the system’s performance, which provide some innovative methods to break through the limitation imposed by the system.