Research On Radio Resource Management And Optimization Technologies For Millimeter Wave Networks

Continuously pursuing higher data rates,higher reliability and lower latency is the evolution direction of wireless communication systems.With the increasing number of user equipments,and the popularity of a series of applications requiring ultra-high-speed data transmission,such as ultra-high definition(UHD)video,virtual reality(VR),etc.,the spectrum resources in the existing microwave band(e.g.,sub 6 GHz)become very crowded,which have been unable to meet the growth of users' requirements.Fortunately,the millimeter wave(mm Wave)band is rich in spectrum resources and can provide ultra-high data rates.Therefore,mm Wave communication is becoming one of the most concerning technologies in both 5G and the future mm Wave based wireless local area networks(WLANs,e.g.,IEEE 802.11ad/ay).However,although the mm Wave band has a large number of available spectrum resources,it also faces many challenges,such as large path loss,easy to be blocked by obstacles,thus the communication distance is greatly reduced.In order to realize long-distance and high-quality mm Wave communication,beamforming technology can be used to concentrate the energy of transmitting signal and receiving range in a narrow beam to achieve directional transmission.However,the coverage of a narrow beam is limited,to provide good signal coverage,one of the most direct and effective methods is to deploy mm Wave networks densely,whereas some new challenges will be introduced.For example,the management of dense networks,performance optimization,and radio resource allocation will become very complex.Therefore,this dissertation is devoted to solving the problem of performance optimization and radio resource management for dense mm Wave networks.Specifically,this dissertation makes a systematic study from the aspects of random access and beamforming training(BFT)protocol,beam management,network architecture,and intelligent radio resource allocation,etc.The contributions and contents of this dissertation are summarized as follows:Firstly,to solve the problem of high collision probability and low BFT efficiency in association beamforming training(A-BFT)phase of the IEEE 802.11 ad based mm Wave network,enhanced random access and BFT mechanism are proposed.This mechanism includes Separated A-BFT(SA-BFT)and Secondary Backoff A-BFT(SBA-BFT).SA-BFT can not only provide more A-BFT slots to alleviate collision,but also maintain the compatibility with IEEE 802.11 ad standard when SAB-BFT is introduced.SBA-BFT further reduces the collision probability and improves the utilization of A-BFT slots by performing a second backoff.The mechanism is modeled as a three-dimensional Markov chain,and the performance analysis shows that the proposed mechanism can greatly reduce the collision probability in the A-BFT phase and significantly improve the efficiency of random access and BFT.Secondly,in mm Wave based unmanned aerial vehicles(UAV)Mesh networks,many challenges are brought by the collective working of UAVs and the beam misalignment of the directional communication links.Therefore,this dissertation studies how to ensure the robustness of the mm Wave UAV Mesh network from the perspective of beam management and network self-healing.Specifically,one of the solutions for the beam misalignment between UAVs is fast beam tracking mechanism.Considering the fact that,there is a group of UAVs flying in the sky,the link failure is easy to happen when some UAVs become invalid,which affects the network connectivity and data transmission.Thus,this dissertation proposes a self-healing mechanism to find an alternative link that just bypasses the invalid UAV(s).This ensures the anti-destroy abilities of the mm Wave UAV Mesh network.To solve the challenges brought by dynamic changes of UAV group leader during flight,an efficient UAV group leader re-selection mechanism is proposed.It can reduce the overhead of BFT when the ground base station performs UAV group leader switching.Thirdly,the directional transmission and dense network deployment of the mm Wave network will make it difficult to carry out beam management and interference coordination,which will affect the overall throughput of the network.Therefore,this dissertation proposes to use deep neural network(DNN)to perform beam management and interference coordination.In this scheme,the problem of beam management and interference coordination is modeled as an optimization problem of radio resource allocation.By designing a heuristic algorithm,the training data of the DNN can be generated.Moreover,the designed algorithm is also used as a benchmark to compare the deep learning method.Research results verify the feasibility and advantages of the proposed deep learning-based radio resource allocation in dense mm Wave network,that is,the deep learning-based method can obtain considerable sum-rate to the conventional radio resource allocation algorithm with relatively less computing time.Lastly,considering that the future WLAN supports both the low-frequency microwave band and the high-frequency mm Wave band,it is difficult to achieve efficient network management and cooperation by using the distributed network architecture.Therefore,a control plane and data plane decoupled WLAN architecture is proposed.In the proposed network architecture,an innovative multi-beam transmission scheme and a high frequency and low frequency cooperation scheme are proposed to improve users' throughput and communication robustness.Research results show that the throughput of users is greatly improved by using the multi-beam transmission mechanism.In addition,the multi-beam transmission and the high frequency and low frequency cooperation mechanism significantly improve the robustness of mm Wave communication.The above research work involves the design of mm Wave network architecture,the improvement of network management and communication protocol,the optimization of radio resource allocation,etc.These research ideas are closely related to each other.They jointly build high-speed,high reliability,and low delay mm Wave communication systems,which may provide a theoretical foundation and technical guidance for the application of mm Wave communication technology in the future.