Research On Joint Detection Algorithm Based On Factor Graph In MIMO-LDPC System

Multiple-Input Multiple-Output(MIMO)technology has become a research hot topic in the field of wireless communication in recent years because of its high spectrum utilization.In order to meet the demand of higher communication bandwidth,it is inevitable to increase the number of antennas,which leads higher requirements on the performance and complexity of the signal detection algorithm of the MIMO system.LDPC(Low Density Parity Check Codes)can effectively improve system reliability as a superior ability to resist random error channel coding.Therefore,the joint-detection and decoding scheme has important research value for MIMO and LDPC systems.The belief propagation algorithm based on factor graph is a new method to solve the communication problem in recent years.The MIMO signal detection problem can be abstracted into a probabilistic reasoning problem.There will be a probability dependency between adjacent nodes in this graph model.Therefore,the method of belief propagation can be used to complete the probabilistic reasoning and update process of nodes in the graph,which reduces the complexity of the algorithm.This dissertation focuses on the research of the MIMO-LDPC system.The joint detection algorithm of MIMO signal and LDPC based on factor graph has been researched.Firstly,the background of MIMO technology is reviewed.The principle of traditional MIMO signal detection algorithm is introduced.The performance of each is analyzed by simulation.Then,we introduce the principle of factor graph modeling and BP algorithm,and adopt the characteristics of BP algorithm for MIMO signal detection and LDPC decoding.We combine the signal detector with the variable nodes in the decoder factor graph.Based on the factor graph MIMO-LDPC joint iterative detection algorithm,the algorithm can achieve better performance than the individually detection and decoding system.For the problem that BP algorithm has poor performance in high-order modulation MIMO system,a non-binary BP(Non-Binary Belief Propagation,NB-BP)algorithm is proposed,which realizes high-order modulation by converting between symbol probability and bit probability at variable nodes.The MIMO-LDPC joint iterative detection algorithm based on factor graph in the system is simulated and analyzed,which demonstrated that the joint detection system can achieve a performance gain of 1 dB with an operational complexity two orders of magnitude lower than the MMSE-PIC system.