Beam Selection Algorithm For Millimeter-wave Massive MIMO

Millimeter-wave massive MIMO and dense heterogeneous networks are considered as the critical enabling technologies for 5G communications,and can greatly improve network capacity and transmission rate.However,the UE beam selection and inter-cell interference are the main problems and technical challenges for heterogeneous networks with densely deployed millimeter-wave small cells.Therefore,this thesis mainly completes the following researches:1.Due to the popular beam selection algorithms have a large amount of information feedback and a complicated calculation process,and are difficult to be applied to the millimeter-wave heterogeneous networks with densely deployed small cells,a position fingerprint-based millimeter-wave multi-cell beam selection scheme is researched.The researched scheme can be divided into two stages: offline information collection and online UE beam selection.In the offline phase,a series of fingerprints are first extracted for the coverage area of each millimeter-wave small cell,and the beam information of each fingerprint is determined.Then,the beam information corresponding to the fingerprints is saved in the specific fingerprint data block,and the fingerprint database can be obtained.During the online phase,the UE periodically performs the current position fingerprint measurement and reporting process.Then,the controller can determine the optimal beam and the strongest interference beams for the UE by position fingerprint matching and distinguish between the best beam and the strongest interference beams to suppress inter-cell beam interference.The researched scheme effectively avoids a large amount of information feedback and calculation measurement of the UE.The simulation shows that the proposed method of inter-cell beam interference suppression can effectively improve the SINR of the signal received by the UE and improve the system sum-rate.2.In order to further improve the timeliness of beam selection in millimeter-wave heterogeneous networks with densely deployed small cells and avoid a large amount of information feedback and measurement process,a BDS-based millimeter-wave massive MIMO beam selection algorithm is researched.To meet the beam selection process,a BDS is proposed and designed,and each beam transmitted by the millimeter-wave AP is identified by a specific BDS,and the best beam for the UE is then determined by detecting the BDS.In addition,the beam interference among adjacent cells is reduced through the inter-cell cooperation.The simulation results show that the researched algorithm can effectively avoid the computational complexity caused by exhaustive search,simplify the beam selection process,and improve the timeliness of beam selection.Especially in the millimeter-wave heterogeneous networks scenario,the BDS-based multi-cell cooperative beam selection algorithm can effectively suppress inter-cell beam interference,and the system sum-rate can be further improved.