Research On Transmission Optimization Design For RIS-empowered Mmwave Mobility Enhancement

Millimeter Wave(MmWave)will be used in 5G and beyond with the continuous pursuit of its mobility enhancement.However,due to the large path loss and blocking affected,high hardware cost and power consumption,it is obviously unrealistic to rely on dense deployment of mm Wave base stations to ensure the performance of the mobility-enhancement system.Reconfigurable Intelligent Surface(RIS),an emerging information-material technology,is attracting increasing attention in recent years for its significant capability to manipulate the radio propagation environment with low hardware cost and power consumption.The widespread deployment of RIS in millimeter wave systems is the key to improve radio propagation environment in an on-demand fashion.In order to fully exploit the performance potential of RIS-mmWave system,it is necessary to obtain perfect channel state information(CSI)and jointly optimize active beamforming of base station(BS)and passive beamforming of RIS.Further consider the algorithm complexity reduction of key transmission methods to achieve efficient and reliable RIS-mmWave system transmission.However,since RIS itself is usually equipped with a large number of elements without signal processing capabilities,which poses a huge challenge to key transmission methods such as beam training,channel estimation,and multi-user transmission for RIS-mmWave system.At that point,this thesis research on key transmission methods such as beam training,channel estimation,and beamforming for RIS-mmWave mobility enhancement system.The main research achievements of this thesis include:(1)Noting the intrinsic connection between RIS-mmWave beam management and effective channel acquisition,a novel machine learning-assisted efficient beam training and channel estimation method is proposed for RIS-mm Wave systems.Specifically,a novel semi-supervised learning model is designed to achieve fast online beam training by exploiting users’ location information firstly,which obtains coarse angular domain information directly without estimation to drive refined channel estimation.Secondly,an angular domain information(ADI)-assisted compressed sensing cascaded channel estimation algorithm is proposed to estimate the RIS-dependent cascade channel,which uses the coarse angular domain information predicted by the learning model to drive the block orthogonal matching pursuit algorithm for channel estimation.Then simulation results show that the performance of the proposed beam training and channel estimation method outperforms the reference scheme in terms of system overhead and channel estimation error.(2)To effectively realize efficient multi-user transmission,a transmission strategy based on beam-domain non-orthogonal multiple access(NOMA)is proposed for RIS-mmWave multi-user system.Firstly,a beam-domain NOMA multi-user system with intra-cluster NOMA and inter-cluster time division multiple access(TDMA)is proposed,which combining NOMA with RIS makes full use of space resources.At the same time,considering the mobility-enhanced RIS-mmWave multi-user system,based on gaussian process for machine learning(GPML)and gaussian mixture model(GMM),a dynamic user clustering algorithm is proposed for beam-domain NOMA multi-user system.On this basis,a beamforming and power control scheme is proposed for beam-domain NOMA,and the RIS reflection phase is adjusted to make it flexibly match different user clusters.At last simulation results show that the proposed multi-user transmission method has a significant improvement in sum rate performance compared with K-means-based user clustering algorithm and TDMA system.