Research On Hybrid Precoding For Millimeter Wave MIMO Communication Systems

The larger available bandwidth of millimeter wave(mmWave)communications and higher spectral efficiency of massive MIMO technology can bring almost an order of magnitude increase in system throughput,making mmWave massive MIMO communications one of the most promising key technologies in 5G communication systems.However,the application of large-scale antenna arrays has greatly increased the hardware implementation cost,energy consumption,and computational complexity of the system.Therefore,how to design low complexity precoding algorithms with satisfactory performance under affordable overall cost is challengable.In this paper,the hybrid precoding algorithms for multiuser mmWave massive MIMO systems are investigated.Our work mainly includes the following three parts:1)The mmWave channel and mmWave massive MIMO system architecture are studied.Firstly,based on the mmWave channel characteristics,such as large path loss and finite scattering,the modeling of narrowband and wideband mmWave channel is investigated respectively.Then the hybrid precoding architecture dedicated for mmWave massive MIMO system is briefly addressed,where the partially connected structure and the fully connected structure are compared and analyzed.2)The hybrid precoding is designed based on the narrowband mmWave massive MIMO system.In the beam training stage,an interlaced scaning(IS)-based scheme is first proposed to reduce the beam training overhead,and then a selection probability(SP)-based scheme is proposed to improve the flexibility of beam training.In order to avoid beam conflict among different users,the beam allocation problem is formulated as a multi-objective bilevel optimization problem.Under the premise that the number of the quality of service(QoS)-satisfied users without beam conflict is maximized,we aim to maximize the equivalent channel gain of the QoS users.A QoS-constrained multiuser beam allocation algorithm is then proposed to solve the beam conflict problem,which can effectively improve the spectrum efficiency of the system.3)Based on the wideband mmWave massive MIMO system,a low complexity hybrid precoding scheme is proposed,where we aim to maximize the spectral efficiency of the multiuser mmWave massive MIMO system.By introducing the interference-free spectral efficiency as a bridge,we decouple the analog precoding and digital precoding design,which can substantially reduce the computational complexity.In the stage of analog precoding design,in order to avoid beam conflict and maximize the interference-free spectral efficiency of the system,a beam allocation method based on the Hungarian algorithm is proposed.In the stage of digital precoding design,the equivalent channel matrix of each subcarrier is diagonalized to eliminate multiuser interference.Finally,the effectiveness of the scheme is verified by simulation.