Research On Message Passing Detection For Multi-user Massive SM-MIMO System

Massive Multiple-Input Multiple-Output(MIMO)is a key technology of the Fifth Generation Mobile Communication System.Due to the small number of antennas,traditional MIMO is difficult to meet the needs of future wireless communication.In massive MIMO system,the base station is equipped with dozens or even hundreds of antennas,which greatly improves the spectral efficiency and channel capacity of the system.In order to maintain its key advantages and alleviate the limitations of practical application,Spatial Modulation(SM)has gradually become a new research hotspot.It divides the transmitted bits into two parts,one for symbol modulation and the other for selecting the transmission antenna.Therefore,SM can not only reduce cost of multi antenna technology,but also ensure good data transmission rate.In multi-user massive Spatial Modulation MIMO(SM-MIMO)system,users use SM to send information,while the base station is equipped with large-scale receiving antenna to receive information.Under the same spectral efficiency,multi-user SM-MIMO can obtain more performance gain than traditional multi-user MIMO.This paper mainly studies multi-user massive SM-MIMO system.Considering reducing the complexity of detection algorithm and improving system performance,the specific research contents are as follows:(1)Message Passing Detection(MPD)is a detection algorithm based on belief propagation,which is suitable to be employed in multi-user massive SM-MIMO systems.However,the complexity of traditional MPD algorithm is still high.To solve this problem,this paper investigates the MPD algorithm based on edge dynamic scheduling.According to the reliability of observation and variable nodes,The Reliability of Observation Nodes(TRON)-MPD and The Reliability of Observation and Variable Nodes(TROVN)-MPD dynamic scheduling schemes are proposed.During the iterative detection,to reduce the complexity,the information of edges is updated based on the reliability of the edges.Theoretical analysis and simulation results show that,compared with the traditional MPD algorithm,both the TRON-MPD and TROVN-MPD schemes can sharply decrease the algorithm complexity at the expense of some bit error rate performance losses.Meanwhile,the better tradeoff can be achieved by setting proper threshold values in the proposed dynamic scheduling schemes.(2)For multi-user SM-MIMO system,based on the MPD detection algorithm,the Layered MPD(LMPD)algorithm which based on layered message passing mechanism is introduced to speed up the convergence speed of the algorithm and reduce the overall computational complexity when the algorithm reaches the optimal performance.Secondly,the performance of MPD algorithm is not particularly superior.Considering the advantages of error correction code,Low Density Parity Check(LDPC)is combined with SM-MIMO system,and a joint LMPD-Belief Propogation(JLMPD-BP)algorithm is further proposed.LMPD algorithm can use the probability information fed back by BP algorithm to make the information transmission in the overall algorithm more reliable,and then improve the system performance.Theoretical analysis and simulation results show that compared with the traditional MPD algorithm,LMPD algorithm can accelerate the convergence speed of the algorithm without losing the performance of Error Bit Rate;At the same time,the system gives full play to the advantages of LDPC code,and the system performance has been greatly improved.Compared with LMPD-BP algorithm which is equivalent to two independent detection units without feedback,the proposed JLMPD-BP algorithm can further improve the system performance.