An optimum two-stage partial parallel interference canceller for CDMA systems

The parallel interference cancellation (PIC) is considered as an effective yet simple detector for multiuser communications in direct-sequence code division multiple access (DS-CDMA) systems. However, the system performance may be deteriorated due to unreliable interference estimate in early canceling stages. Thus the partial cancellation factor (PCF) is introduced to control the interference canceling level resulting in the partial PIC approach. While the PCF is critical, the existing methods for determining optimum PCF is either not general or not accurate. In this paper, we propose a method deriving a close-form solution for the optimum PCF in a two-stage partial PIC receiver. Our results apply to either power balanced or unbalanced, synchronous or asynchronous systems. Simulation results show that our theoretical optimal PCF values are closed to the empirical ones. Also, the optimal two-stage partial PIC even outperforms three-stage full PIC.