Over the past few decades, code-division multiple-access (CDMA) systems have been widely used for military and commercial applications. However, the multiple-access interference (MAI) greatly degrades the system performance. To overcome this drawback, many researchers have put forward multiuser detectors (MUDs) to handle the MAI properly. The MUDs make use of the MAI to help make accurate decisions, which is unlike the conventional detectors. In this project, existing multiuser detectors are studied thoroughly. This dissertation first presents several typical multiuser detectors and gives the analysis of system performance. In some linear detectors, inversion of correlation matrix is not welcomed in practice. Interference cancellation schemes in parallel or serial structures are also studied. Then both blind and non-blind adaptive detection algorithms are discussed. The blind multiuser detectors are very welcomed because they do not need training sequence and the a-priori knowledge of interferential users. The simulation results show that all these multiuser detections can achieve a better system capacity and a lower bit error rate, however, different receivers can gain its best performance under different environments. At last, several practical problems and future research topics are presented. With the multiuser detection technique becoming increasingly important, their applications will also be expanded continuously.

Master of Science (Communications Engineering)