Multiuser Detection Algorithm In Mobile Communication

There exists serious multiple access interference in CDMA communication system, because it'S very difficult to keep orth between spread codes, Which limited system capacity. Multiple access is considered has different statistic character with white noise in multiuser detection, so multiple access could be suppressed by union detection. The main works of the paper can be summarized as follows: I. A T-H Neural Network-based multiuser detection algorithm is proposed, which has rapid operation speed and globally stable point. 2. A Variable Weight General Energy Function is proposed and apply into subspace multiuser detection, which exact parallel multiple principal components quickly and has lower computation complexity. 3. Input correlation matrix without expected user signal is obtained after expected user signal is filtered by array antenna, which avoid signals loss resulted by at original period. 4. A support vector machine-based multiuser detection algorithm is proposed support vector machine consider experience risk and capacity control synthetically has more powerful small sample generalization by introduce Construct Risk Minimization. 5. Cascade spatio-temporal processing and joint-domain spatio-temporal processing models is proposed by introduced array signal processing. This paper is supported by National ?63?plan.