Research On Subspace-based Blind Adaptive Multiuser Detection In Thier-generation Mobile Communication System

Code-division multiple-access (CDMA) is an air interface standard of the third generation mobile communication system. The spread-spectrum code of CDMA isn't orthogonal strictly in a general way because the connection of the stochastic multi-user. The nonzero cross-correlation coefficient will cause multi-access interference (MAI). The MAI increases the error rate and limits the capability of the systems. The Multi-user Detection (MUD) eliminates the MAI to improve the performance of the receivers, increase the capability of the system and overcome the infection of the "near-far" effect using the inherent structure information of condition code of the CDMA users. MUD is one of the key techniques of the third generation mobile communication system. Moreover, the blind MUD technologies are of special interest in the mitigation of MAI in the CDMA systems as only the a priori knowledge of signature waveform and the timing of the desired user are needed. The subspace blind MUD based on the tracking and estimate of the signal subspace with lower computational complexity and better convergence performance.The purpose of this paper is to research subspace-based blind adaptive multiuser detection. The main works of this paper can be summarized as follows:1. The classic algorithms of the rank estimate and subspace tracking modules in the subspace multi-user detectors are analyzed theoretically and researched in the emulation experiments. The improved orthogonal projection approximation subspace tracking (OPAST) algorithm is proposed. The improved algorithm offers the better convergence performance and tracking effect of signal subspace eigenvectors as demonstrated by computer simulations.2. Compared with the Least Mean-Square (LMS) and Recursive Least Square (RLS), the Kalman filtering blind adaptive MUD demonstrates faster convergence speed and lower steady-state excess output energy in apaptation. But it has higher complexity. To reduce the computational and improve the performance of the algorithm, this paper proposes a modified Kalman filtering algorithm integrating the subspace concept and improved OPAST algorithm.3. The simulations of the proposed subspace-based Kalman filtering blind adaptive MUD algorithm are operated in an Additive White Gaussian Noise (AWGN) channel, a slowly time-varying Rayleigh-fading channel, a multi-path channel and a dynamic environment with variable number of users. Compared with the original algorithm, the improved one has lower computational complexity, faster convergence speed and better performance of MAI suppression.