Study On Iterative Detection Algorithm Based On LDPC Codes In MIMO System

The future wireless communications systems should be able to provide reliable high data ratetransmissions at low cost.It is believed that the wireless MIMO(Multiple Input MultipleOutput)technology ,iterative detection and channel coding will become some of the key solutions tothe beyond third generation (B3G) or the fourth generation (4G) mobile communicationssystem.MIMO technology can offer an improved capacity through using multiple antenna. Theiterative detection technology includes soft-input soft-output (SISO) detection and channel codingtechnology.It can exploit the information passed by SISO decoder to suppress interference and has agood performance, thus has gained much attention from the researchers. The main task of thisdissertation is to do some researches on iterative detection technology and the main structure issummarized as follows:1) First of all,introducing the decoding principle of low-density parity-check codes(LDPC) .Then based on LDPC codes,the structure of an iterative detection system is proposed.2) Based on research on traditional SISO parallel Soft Interference Cancellation plus MinimumMean Square Error (SIC-MMSE) detection algorithm ,one algorithm based on symbol feedbackafter its soft decision and another algorithm based on ordered reliability are analysed in detail.Thelater two algorithm can be implemented with a larger time delay than the SISO parallel detectionalgorithm, but with a better performance.3) Researching on the SISO layer iterative detection algorithm in iterative detection systemusing LDPC codes,some improved algorithms are proposed based on the idea of suboptimal orderand grouping. The improved algorithms perform better through reducing error spreading andobtaining diversity gains.4) The performance of detection algorithm based on MMSE formula under imperfect channelestimation is analysed. By considering statistic information of channel estimation error, the softoutput information of the iterative algorithm is modified under estimated channel information.