Research On Multi-user Detection For MC-DS/CDMA Systems

Multicarrier DS/CDMA system is extensively studied recently, which can advance the transmission rate of wireless communication systems, achieve high spectrum efficiency and combat frequency-select fading efficiently compared to single-carrier DS/CDMA system. It adapt to high data rate transmission greatly. However, the loss of orthogonality of the signature spreading codes after the transmission leads to the multi-access interference(MAI), which will cause the degradation of the performance of the system. Multiuser detection (MUD) is one of the most powerful solutions to cancel the MAI. The MLSD proposed by S.Verdu can get the minimum probability of error bit, but its computational complexity is increased exponentially when the number of the users is increased, It can't fulfil the need of real-time communication.In order to make implementation of MLSD become practical, the main works of communication signal processing focuses on sub-optimum multiuser detection technique, which compromise the performance with computational complexity. Discrete particle swarm optimization(DPSO) is a new optimization technique, which can be implemented with ease and few parameters need to be tuned. It has been successfully applied in many areas. In this paper, We applied the DPSO algorithm to solve the MUD problems and construct multiuser detector(DPSO-MUD), using combinatorial optimization ability of DPSO to determine special combination of each user symbol bit transmitted. On the other hand, a new modified DPSO algorithm(SADPSO) which involved the Simulated Annealing strategy is proposed, in order to improve the convergence speed and avoid plunging into the local minimum. Then construct new multiuser detector(SADPSO-MUD). Simulation results show that the two given Detectors have greatly improved in terms of bit-error-rate , near-far resisrance, and SADPSO-MUD have great advantage of convergence property over DPSO-MUD. Furthermore, the MC-DS/CDMA system has better performance than DS/CDMA system under the same bandwidth and the same algorithm.