Study On Iterative Decoding Technique In MIMO System

With the growth of communication requirements, it is a huge challenge that how to use limited spectrum resource to achieve high-rate, reliable wireless data transmission. Due to adopt the multi-antenna technology, MIMO systems is an effective solution of high-rate and high reliable wireless transmission. The research of MIMO technology mainly includes the design of space-time coding and detection of MIMO signals. In the aspect of MIMO signal detection, in the MIMO systems that concatenated with channel coding, using the Turbo principle, we can make the soft detection of MIMO signals and approach the performance of Maximum Likelihood reception. It is a key issue to design high-performance, low-complexity soft detector of MIMO signals in the aspect of MIMO signals detection.Because of this, the thesis studied on the technology of iterative decoding in MIMO systems, and tried to get some improvement on the performance and complexity of the iterative decoding algorithm. The main content is displayed as follows.1) Starting from the propagation characteristics of radio channel and the expressions of MIMO systems, analyzed the capacity of MIMO channel and the trade-off criteria of Diversity gain via Multiplexing gain. By introduced the criteria of space-time coding in MIMO systems, analyzed the principle of two typical codes in space-time coding: Alamouti code and V-BLAST architecture. Finally, the simulations of these two codes were given.2) Studied on the optimal MAP soft detector and suboptimal LSD-MAP soft detector on the basis of structure of MIMO systems concatenated with channel coding. Mainly analyzed the principle of sphere decoding algorithm in LSD-MAP, and combined with the characteristics of modulation signals, the searching range was decreased to reduce the complexity. Finally, given the fundamental steps of algorithm to realize the iterative decoding of coded MIMO system.3) In order to further reduce the complexity of LSD-MAP algorithm, the FP-MAP algorithm, which used the soft information during the searching process of sphere decoding, was researched. Theoretical analysis showed that, compared with the LSD-MAP algorithm, FP-MAP algorithm had relatively lower complexity. The simulation results also showed that the BER performance of FP-MAP algorithm was equal to the LSD-MAP algorithm. 4) Proposed the new iterative decoding algorithms with lower complexity for Linear Dispersion Codes(LDCs). First by the special coding structure of LDCs, obtained the equivalent transmission model. After, on the basis of this, using the thought of interference cancellation, designed the Vector Gaussian Approximation, Scale Gaussian Approximation and Minimum Mean Square Error algorithms for the soft detection of LDCs. Theoretical analysis and simulation showed that, compared with the LSD-MAP algorithm, the methods this thesis designed had lower complexity and the performance still met the application requirement. The methods realized the effective trade-off between performance and complexity.