Performance of a GA-Based Interference Canceller for MC-CDMA Communication Systems Over Frequency Selective Fading Channels

In this paper, we investigate the performance of multistage GA-based parallel interference cancellation (PIC) multiuser detectors for multi-carrier direct-sequence code-division multiple-access (MC-CDMA) communication systems over frequency selective fading channels. The genetic algorithm (GA), which tries to reduce the complexity in optimal detection due to the greedy exhausted searching, performs the adequate sub-optimal multiuser detection for CDMA systems. Therefore, in this paper, we further investigate the GA-based PIC multiuser detection for MC-CDMA systems. Simulation results show that with a lower generation of G_n=5, 10, the constant weight PIC scheme outperforms the conventional PIC (CPIC) especially at heavy system load. However, with higher generation G_n=20 in GA, the conventional PIC (CPIC) scheme contrarily outperforms the constant weight PIC for all system load.