Research On Precoding And User Scheduling In Millimeter-Wave Massive MIMO System

With the rapid development of multimedia services,the demand for wireless high-speed data communication is ever increasing.In order to deal with the future explosive growth in mobile data,the fifth generation of mobile communication(5G)system come into being.As one of the key technologies for 5G system,Massive MIMO has attracted much attention as it can greatly improve the spectral efficiency of the system.Moreover,the millimeter-wave has become a research hotspot in recent years due to its abundant available spectrum resources and easy integration of ultra-large antenna arrays."Therefore,this paper focus on the millimeter-wave Massive MIMO system and its related technologies.In millimeter-wave Massive MIMO system,Massive MIMO technology can be combined with millimeter-wave by using hybrid beamforming,so as to form directional narrow beams with high gain to overcome the impact of large path loss in millimeter-wave high-frequency band.Different from the traditional full digital precoding technology,hybrid beamforming technology includes both analog and digital domain precoding.By transferring part of precoding processing to analog domain,the hybrid precoding can greatly reduce the complexity of the system.Firstly,based on the hybrid beamforming technology of equivalent channel,this paper mainly focuses on the beam selection problems in analog domain.On the other hand,due to the limitation of the number of radio frequency links at the base station and the influence of inter-user interference,the base station often cannot serve all candidate users on the same time-frequency resource.Therefore,it is critical to adopt a reasonable user scheduling algorithms.Most of the existing user scheduling algorithms in the millimeter-wave Massive MIMO system are discussed in the scenario of single-cell,multi-user and users are configured single antenna.However,in actual millimeter wave MIMO system,the number of cel:ls is much larger than one and users often have multiple antennas.Therefore,this paper will further study the user scheduling problem in a more complex millimeter-wave scenario that configures multi-cell,multi-user and base stations and users are equipped with multiple antennas.The main work of this paper is as follows:1.In the hybrid beamforming technology based on the idea of equivalent channel,the analog domain beam selection problem in millimeter-wave Massive MIMO system is studied.Firstly,two beam selection algorithms based on exhaustive search mechanism are introduced.Because of the shortcomings of these two algorithms,a new hybrid beam selection algorithm is proposed.This algorithm allows users to select the appropriate beam selection criteria intelligently according to the maximum receiving power.The simulation results show that the new hybrid beam selection algorithm can achieve a good compromise between the beam selection delay and the system SINR performance by using the appropriate RSRP threshold.2.To solve the problem of high search complexity of exhaustive search algorithm,a threshold search algorithm with low complexity is proposed.This algorithm introduces the concept of performance threshold based on the exhaustive beam search algorithm,and achieves the selection of beam pairs by comparing the performance value and performance threshold of each analog domain beam pair.The simulation results show that the threshold search algorithm is superior to the exhaustive search algorithm.By setting an optimal performance threshold,the threshold search algorithm can greatly reduce the search complexity while ensuring the system performance.3.The user scheduling problem is studied under the complex scenario of millimeter wave multi-cell multi-user and multi-antenna configuration for both base station and user.Three scheduling algorithms are simulated and compared in two 5G scenarios of the uplink and downlink modes.The simulation results show that the spectrum efficiency of downlink mode is better than that of uplink mode in both two scenarios.In each link mode,the proportional fair single-user scheduling algorithm is better than the random beam single-user scheduling algorithm,but inferior to the proportional fair multi-user scheduling algorithm.And with the increase of the number of scheduling users,the performance of proportional fairness.multiuser scheduling algorithm is better,but the growth rate slows down gradually.