Research On Signal Detection Algotithms For Massive MIMO Systems

In today's society,with the rapid development of wireless communication technology,more and more intelligent terminal devices are connected to wireless network.People are gradually stepping into the 5G era,and the demands for communication systems are becoming higher.In order to meet the requirements of future communication technologies in terms of node density,communication rate and spectrum utilization,massive MIMO technology,as a key technology of 5G,has shown an important role and great potential,and has become one of the research hotspots.For large-scale MIMO,signal detection determines the performance of MIMO system.However,affected by the increase of antenna array,the channel environment becomes extremely complex,which brings challenges to the accuracy and complexity of signal detection at the receiving end.In order to solve the main problems of high computational complexity and poor detection performance,this paper focuses on the corresponding superior detection algorithm.The main work is as follows:1 This paper first introduces the signal model of MIMO system,then analyzes and compares several signal detection algorithms.MATLAB simulation software is used to analyze and compare the bit error rate of the proposed algorithms in different environments,and their computational complexity is discussed.2 Based on the Chebyshev iteration(CI),a low complexity and high performance coupled Chebyshev Iteration(CCI)detection algorithm is proposed,which achieves the BER performance closer to MMSE than the traditional linear detection algorithm.CCI algorithm introduces a direction vector to improve the parallelism of relevant vector calculation in the detection process,and does not need to solve the complex large-scale matrix inverse operation involved in MMSE.At the same time,it reduces the multiplication calculation of matrix and matrix and matrix and vector,and speeds up the iterative process,thus greatly reduces the complexity.3 Signal detection model is converted into a convex optimization problem.By convex optimization method of alternating direction fast convergence properties of multiplier method(ADMM),the initial solutions are obtained,Then the search algorithm(NSA)is improved on each element in the field of vector search optimization,which avoids the high complexity of the search process,Combining the ADMM algorithm to propose a high convergence rate and high detection performance algorithm N-ADMM,the better bit error rate performance is obtained while keeping the complexity at a lower value.