The Research Of Detection For Massive GSSK-MIMO Systems

Massive Multiple Input Multiple Output(MIMO)systems with tens or hundreds of antennas at both transmitter and receiver sides is capable of achieving that a large number of antennas send data at same times,full use of space resources,can greatly improve the spectrum efficiency and power efficiency and link reliability.As one of the key technologies of the next generation communication system,large-scale MIMO system is widely concerned by industry and academia because of its potential to be underestimated.However,challenges and opportunities coexist.With the increase of the number of transceiver antennas,the complexity of signal detection is increased greatly.Therefore,how to obtain higher detection performance with lower complexity is an urgent problem to be solved at present,which is this paper focus on.This study aims to research the detection algorithm for massive MIMO systems and two effective detection algorithms are proposed.The main contents and innovations of this thesis are as follows:Firstly,the massive MIMO system with mathematical model is given,and some traditional detection algorithms for massive MIMO systems are analyzed,including Maximum Likelihood(ML)algorithm,Matched Filter(MF)detection,Zero forcing(ZF)detection,Minimum Mean Square Error(MMSE)detection and the detection based on Successive Interference Cancellation(SIC)is analyzed.Secondly,a low-complexity detection based on the MMSE detection for MIMO systems in this thesis.Using on the MMSE detection result as the starting point,the proposed scheme searches the constellation subspace using the ML criterion,which can improve the detection performance.Compared with the ML detection algorithm,the proposed algorithm does not need to search the whole constellation space,so it can obtain the detection performance near ML by searching a limited range of subspace without introducing substantial computational complexities.Thirdly,an improved algorithm of MMSE based on compressed sensing(CS)is proposed in this thesis.Considering the sparsity of residual error between transmit signal and the MMSE detection results,we can effectively estimate the position of the error symbol in MMSE detection results by using the theory of CS,and correct it to improve the detection performance,which needs less computational complexities than the proposed low-complexity detection based on the MMSE detection.The Two proposed detection algorithms are simulated.The simulation results show that the proposed algorithms can achieve efficient performance.So the study proposes a new way for the publication and research for massive MIMO systems.