Performance Analysis Of The PIC Multi-User Detection Algorithm For CDMA System

Multiple access interference (MAI) is a major factor which limits the capacity and the performance of CDMA systems. However MAI is not the inherent problem of CDMA systems, it is because the conventional CDMA systems follow a single-user detection strategy. In multi-user detection, the useful information of multiple users is jointly used to better detect each individual user. In addition to mitigating MAI, multi-user detection also can improve the system capacity and its performance. Therefore, multi-user detection is a very promising technology of mitigating MAI for CDMA systems.Firstly, the principle of spread spectrum communication is discussed. It includes the classification and performance indicators. Then the paper states the principle and classification of multi-user detection in CDMA systems, analyzes and compares the two main categories of suboptimal detectors, linear multi-user detectors and interference cancellation multi-user detectors. In linear multi-user detectors,a linear mapping is applied to the outputs of the conventional detector to produce a new set of outputs, which can provide better performance. Besides, in interference cancellation detectors, estimates of the interference are generated and subtracted out for each user in parallel, which can significantly improve the system performance.Secondly, this paper discusses the principle and performance of the conventional detector and the parallel interference cancellation (PIC) detector separately. Since the conventional detector take MAI only as the background noise without considering the existence of the other users, the existence of MAI has a significant impact on the performance of CDMA systems. Each stage of PIC detector utilizes the data estimates of the initial or previous stage, produces a new set of MAI estimates, and then subtracts out all of the MAI for each user in parallel. Through Monte Carlo simulation in different scenarios, the results illustrate the influence of MAI in conventional CDMA systems. The experimental results show PIC detector can effectively eliminate MAI and improve the system BER.Finally, the PIC multi-user detection algorithm based on the polynomial expansion is proposed, this method can conquer the limitation of error propagation. Aiming at improving the detection reliability, this improved algorithm takes the output of polynomial expansion (PE) detector as its input the data estimates of the first stage of PIC detector. Through computer simulation, the experimental results illustrate the good performance of PE detector. Compared with the conventional detector, simulation results show that the significant performance gains of the improved algorithm can be achieved in the system. The improved algorithm is proved performing better than the PIC detector, decorrelating detector and MMSE detector.