Research On Signal Detection Algorithms For MIMO/Massive MIMO System

As one of the core technologies of 4G,MIMO technology has gradually failed to meet the increasing data rate requirements of people.In recent years,massive MIMO technology has attracted widespread attention from researchers at home and abroad.Massive MIMO technology is a key technology of 5G by deploying a large number of antennas at the base station to form a multi-antenna array,which has the advantages of high spectral efficiency and high energy efficiency,and can significantly improve the system capacity of the wireless communication system.Signal detection at the receiver is a key link in the implementation of the MIMO system.The advantages and disadvantages of the signal detection algorithm will have a significant impact on the performance of the communication system.Therefore,signal detection algorithms with lower design complexity and better performance are especially important for MIMO and massive MIMO systems.The main work and innovations of this paper are as follows:(1)Research on signal detection algorithms in traditional MIMO system.This paper first studies some classical signal detection algorithms in MIMO systems,and simulates their detection performance.Then,the LR algorithm is introduced in detail,and the system model and corresponding signal detection process after LR processing are given.Finally,in view of the high complexity of ML-SIC algorithm in high-order modulation systems,this paper uses LR algorithm to improve the traditional ML-SIC algorithm,and proposes an LR-assisted ML-SIC signal detection algorithm based on ZF criterion.Simulation results and complexity analysis show that compared with the traditional ML-SIC algorithm,the proposed algorithm can obtain near-optimal performance with lower computational complexity.Especially for high-order modulation systems,the performance of the proposed algorithm is superior.(2)The signal detection algorithm for the uplink of multi-user massive MIMO system is studied.Traditional MMSE detection algorithm can achieve near-optimal performance in massive MIMO systems,but it contains high-dimensional matrix inversion with high complexity.To solve this problem,this paper first studies two simplified algorithms of MMSE algorithm such as Neumann series expansion algorithm and Newton iterative algorithm.Then,this paper proposes an improved MMSE signal detection algorithm based on SOR method.In addition,this paper also proposes a region-based initial solution to optimize the traditional zero initial solution.The simulation results and complexity analysis show that the proposed algorithm converges to the MMSE algorithm faster than the Newman series expansion algorithm and the Newton iterative algorithm.In addition,the proposed algorithm converges to the MMSE algorithm faster when using the region-based initial solution than using the zero initial solution,the advantage is especially noticeable when the number of iterations is small.Since the high-dimensional matrix inversion is avoided,the computational complexity of the proposed algorithm is reduced by an order of magnitude compared with the MMSE algorithm.(3)The signal detection algorithm in the downlink of single-user massive SM-MIMO system is studied.This paper first studies the classical Co Sa MP algorithm in CS theory for signal detection in massive SM-MIMO systems.Then,for the problem that the performance of signal detection algorithms based on CS theory is not good in massive SM-MIMO system,this paper proposes a SCo Sa MP algorithm based on SCS theory and gives the specific implementation steps of the algorithm.The simulation results show that the proposed algorithm performs better than the traditional Co Sa MP algorithm under correlated and uncorrelated channel conditions,and can achieve near-optimal detection performance.