CDMA Multiuser Detection Using Unscented Particle Filtering

In this paper, unscented particle filtering (UPF) is applied to multiuser detection (MUD) in synchronous code division multiple access (CDMA) environments. The UPF algorithm consists of a particle filtering (PF) that uses an unscented Kalman filtering (UKF) to generate the importance proposal distributions, which match the true posterior better. The simulation results show that the UPF-based multiuser detector almost always outperforms the PF-based one both in the equal power situation and the near-far case.