| In CDMA systems signals conveying desired information are received in the presence of ambient noise and multiple-access interference (MAI). Therefore multiuser detection, which can effectively inhibit MAI, is the key technique to ensure the system capacity and the quality of wireless communication. Early works on multiuser detection are mostly limited to pure optimization of multiuser detection, giving little attention to the interconnection with other modules. But with the idea of iterative processing being popular as a result of the invention of Turbo code, iterative multiuser detection is receiving more and more attention.The iterative receiver that mainly consists of the SISO (Soft-In Soft-Out) multiuser detector and the SISO channel decoder is considered as a feasible solution to the problem of joint detection and decoding. In this paper we will firstly give an introduction to this iterative structure, in which both the SISO multiuser detector and the SISO channel decoder raise their performance with the information of coded symbols offered by each other until the whole capability approaches to optimum after several iterations.The SISO multiuser detector is the key of iterative multiuser detection, but the optimum SISO multiuser detection has a prohibitive complexity, which leads researchers to paying more efforts to the suboptimum but low-complexity SISO multiuser detection. In this paper we study several low-complexity SISO multiuser detection algorithms, including soft parallel interference cancellation algorithm, soft interference cancellation MMSE algorithm and sensitive bit algorithm, and their performance is compared through system simulation. Furthermore in this paper a soft interference cancellation group algorithm, based on the asymmetric distribution of interference caused by users, is proposed aiming at reducing the complexity of iterative multiuser detection by approximate processing.CDMA2000 1xEV-DO is the data-enhanced system of CDMA2000 1x. In order to increase data transmission rate of its forward link, the modulated symbols are transmitted through 16 parallel code-channels, which are separated by orthogonal codes– Walsh codes. The orthogonality, however, is impaired by multipath fading channels, thus causing multicode interference (MCI) and deteriorating the system's performance at high data transmission rates. In this paper, the low-complexity iterative multiuser detection is introduced into the forward link of 1xEV-DO to eliminate the MCI. The simulation results show that the performance of the iterative multiuser detector is superior to that of the Rake receiver as well as the chip-level equalizer, and hence indicate that the low-complexity iterative multiuser detector is comparatively a good solution to the problem of high-rate data transmission of the forward link of 1xEV-DO. |