Turbo Multiuser Detection Based On Factor Graph In Spread Spectrum Communication System

In the system of spread spectrum communication, for the reason that spread spectrum code sets are not completely orthogonal and the correlations of these code sets are nonzero, it gives rise to the multiple access interference between different users, which severely reduces the reliability in the information transmission. As the Turbo multiuser detection technology has been introduced, we success in solving the problem above. Adopting to the turbo iteration principle, the technology simultaneously processes both multiuser detection and channel decoding, effectively improves the ability of the communication system in avoiding multiple access interference.In this paper, we deal with the problems about Turbo multiuser detection base on the actor graph and sum-product algorithm. Not only we get a detection algorithm with fine performance, but also mix multiuser detection and channel decoding together in a consistent framework, explaining the whole turbo multiuser detection technology in terms of a joint and unified view. The main contents are as follows:1. Describe the principle of turbo multiuser detection detailedly, analyse and simulate two classic algorithm under the turbo multiuser detection model.2. Have a research on the unified model of FG-Turbo multiuser detection in the CDMA system and the BP decoding algorithm of LDPC code. After giving out the extended model of LDPC code, we detailedly analyse the optimum MAP algorithm and simulation performance in the extended model.3. Givie out the unified model of FG-Turbo multiuser detection with faster transmission speed in the MC-CDMA system, and extend it to the model of LDPC code. Research and simulate the soft interference cancellation (SIC) algorithm. According to the interference estimates for soft interference cancellation algorithm, we propose a new soft interference cancellation algorithm based on sensitive bits (SB) called SB-SIC.