Research On Resource Allocation Scheme For UAV-Enabled Wireless Communication Network

In the past few years,the use of unmanned aerial vehicle(UAV)in the civil field has a huge growth,such as air surveillance,traffic control,photography,package delivery,and gradually shows the development trend of close combination with wireless communication.The coming fifth generation mobile communication system(5G)will bring more development space for UAV applications.The high reliability and low delay of 5G will provide high speed networked data transmission channels for low-altitude UAV.The massive internet of things features of 5G can be combined with artificial intelligence technology to lay out the connected ecology of all things,and create a basic environment for intelligent interaction and coordinated flight of UAV.The combination of UAV and cellular mobile communication technology has three main research directions.In detail,UAV work as a temporary cellular base station and provides communication services for ground user.Besides,UAV serves as a relay node to expand the coverage of wireless networks.What is more,UAV acts as a user terminal and performs specified tasks.And this paper also focuses on these three aspects.Firstly,the paper considers the emergency situation that vehicular communication network is destroyed due to natural disasters and the links between vehicles and base station is interrupted.UAV acts as a temporary cellular base station and provides communications services for the vehicles with the help of resources of the adjacent cell.The uplink resource allocation in UAV enabled vehicular communication system is studied.Resource management is based on slow fading parameters and channel statistics rather than instantaneous channel state information in this paper.Device to device(D2D)technology is applied to vehicle to vehicle(V2V)link to ensure reliable communication.Through theoretical analysis,optimal transmission power expressions for cellular users and device to device users are given,respectively.Vehicle to infrastructure(V2I)link selecting resource block and V2 V link multiplexing the resources of V2 I link are formulated as 3-partite graph matching problem.By utilizing k-partite graph matching theory and the Local Ratio algorithm,the 3-partite graph matching problem can be solved.Besides,without considering the selection of resource block,the problem of V2 I link multiplexing V2 V link can be modeled as a bipartite graph matching problem,and Hungarian algorithm can be adopted to solve the problem.What is more,considering the mobility of UAV,greedy algorithms in reinforcement learning is adopted to control the trajectory of UAV,which improves the sum achievable rate of cellular users and D2 D users further.Secondly,the uplink resource allocation problem of cellular communication system when the UAV terminal and the ground users exist at the same time is studied in this paper.The channels between ground base station and UAV are dominated by the line of sight links,which brings both opportunities and challenges.On the one hand,the signal transmitted by the UAV at the ground base station will increase in strength as the path loss is reduced.Simultaneously,a UAV can be associated with multiple ground base stations,which can bring higher macro diversity gain to the system as compared to the ground users.On the other hand,the UAV will also bring strong uplink interference to the ground users.In order to achieve a compromise between the communication performance of UAV and interference control,the weighted sum achievable rates of the UAV and the ground users are maximized by jointly optimizing the base station association and power allocation of the UAV.The optimal associated base station of UAV is derived.Successive convex approximation technology is adopted to solve the power allocation problem considering both UAV and ground users simultaneously,and the upper bound of the objective function was determined by outer polyblock approximation algorithm.Finally,the paper mainly studies the filter design problem at the base station and relay when UAV is worked as the relay node of cellular communication.The optimal objective is to maximize the weighted sum achievable rates of ground mobile users.Due to the problem is nonconvex,which is hard to find an analytical solution.Therefore,the optimized problem is transformed to minimize the weighted sum of mean square error.When solving the transformed problem,the base station filter and relay filters are assumed to be fixed,then the filter at receiver and weight matrix of ground mobile users are determined.Next,the filter matrices at the relay and the base station are solved in turn,and the value are gradually approaching to the optimal solution by continuous iteration.