We employ a multiuser detection (MUD) method using a novel genetic algorithm (GA) based on complementary error function mutation (CEFM) and a differential algorithm (DA) for ultra-wideband (UWB) systems. The proposed MUD method is termed CEFM-GA DA for short. We describe the scheme of CEFM-GA DA

analyze its algorithm

and compare its computational complexity with other MUDs. Simulation results show that a significant performance gain can be achieved by employing the proposed CEFM-GA DA

compared with successive interference cancellation (SIC)

parallel interference cancellation (PIC)

conventional GA

and CEFM-GA without DA

for UWB systems in lognormal fading channel. Moreover

CEFM-GA DA not only reduces computational complexity relative to conventional GA and CEFM-GA without DA

but also improves bit error rate (BER) performance.