Antenna Selection For Massive MIMO

Massive MIMO is an evolution of traditional MIMO.Its key idea is to deploy more antennas in the communication system,usually tens or even hundreds.In general,massive MIMO which uses a large number of antennas can be considered as an enhanced version of traditional MIMO.Therefore,in terms of capacity,efficiency,and reliability,its system performance is significantly better than that of traditional MIMO.However,in practical applications,the main limitation to increasing the number of transmit and receive antennas is usually not the size of the number but the cost of the radio frequency chains connected to these antennas.Antenna selection can achieve a substantial reduction in the number of radio links without significant performance loss,thereby reducing the system's implementation costs.Therefore,this dissertation focuses on the antenna selection in Massive MIMO.Firstly,we concentrate on the characteristics of massive MIMO,as well as the ensuing challenges.We discusse in detail both the theoretical analysis and the measured data.The theoretical analysis shows the system's channel capacity expression.It shows that increasing the number of antennas can increase the channel capacity and reduce the interference between users,and this conclusion is verified by simulation experiments.Then through the analysis of measured data,more features of the system are obtained: The antennas in the array have unbalanced power gains,which illustrate the practicality of the antenna selection.The large number of antennas is associated with high system complexity and strong antenna correlation,which is the main challenge that antenna selection faces in Massive MIMO.Secondly,we focus on the development and improvement of antenna selection.The concepts,advantages,and guidelines of antenna selection are systematically described,as well as various existing classic algorithms.On this basis,the application of antenna selection has been expanded.A time-varying channel model and two more complex MIMO models are discussed,and the corresponding antenna selection model is analyzed.Then,a novel high-performance antenna selection algorithm called minimum condition number algorithm is proposed.The algorithm is based on theoretical analysis of channel capacity,and achieves high performance with the condition number of channel matrix as the breakthrough point.The key idea is to process the antennas one by one,and the selected antenna is required to make the channel matrix corresponding to the new antenna subset have the smallest condition number.Simulation experiments show that the channel capacity of the algorithm can reach very close to the optimal algorithm,which is obviously better than the classical incremental algorithm.With the increase of SNR,this trend is more obvious.Finally,aiming at the challenges brought by Massive MIMO,a novel low-complexity antenna selection algorithm called grouping algorithm based on channel capacity is proposed.The algorithm divides several antennas in adjacent locations into a group,and then selects the antenna with the highest channel gain in each group.The key idea is to avoid selecting the distributed antennas at the same time by grouping,effectively overcoming the correlation between the antennas,and improving the channel state by selecting the antenna with the maximum gain.Because it does not involve time-consuming operations such as permutation and combination,singular value decomposition and loop iteration,but simply requires the gain of the channel,the complexity of the algorithm is very low,so it is suitable for massive MIMO.Simulation experiments show that the algorithm always maintains excellent performance regardless of the correlation between the antennas.With the enhancement of correlation,its performance advantage is more prominent.Therefore,this algorithm has good feasibility.