Research On Efficient Transmission Strategy Of Mobile Edge Computing Network In UAV Patrol Inspection System

Unmanned aerial vehicles(UAVs)have become a popular tool in the field of wireless communications due to their high mobility,flexibility and expandability.UAVs equipped with communication devices can be easily deployed to perform various tasks such as rescue,tracking,etc.Compared to static communication devices,UAVs can move freely in three-dimensional space to establish line-of-sight links with ground users.Especially in areas where communication resources are scarce and the communication requirements of a large number of users cannot be met,UAVs can act as aerial base stations to provide communication services to users.In addition,cellularconnected UAVs are always connected to the ground cellular network,which ensures the flight safety and communication quality of UAVs.Considering the unique advantages of cellular-connected UAVs,they are widely used in various industries.In this thesis,we study two-dimensional scenes and three-dimensional scenes of cellularconnected UAV fixed-point patrols,considering that UAVs equipped with information collection devices will capture target area information at the location-determined cruising points and later process the data in collaboration with ground base stations.The main research content and innovation points of this thesis are summarized as follows.This thesis first investigates the energy minimization problem in a cellularconnected UAV fixed-point patrol scenario.Considering the limited energy of the UAV and the urgency of data processing,this thesis makes full use of the ground network resources and combines mobile edge computing technology for the UAV to compute and offload the data.For the proposed patrol scenario,a mathematical model is first constructed to minimize the energy consumption of the UAV system,which is mainly realized by designing the UAV flight trajectory,computation resource allocation and communication resource allocation schemes.Consider the timeliness of data and the delay sensitivity of monitoring events,it is assumed that the UAV must finish processing the data collected at the previous cruise point before flying to the next cruising point.Therefore,the problem is further divided into two subproblems: optimal transmission strategy between two consecutive cruising points and traversal order design.The former is a nonconvex problem,which is further divided into two subproblems,and then solved by an alternate optimization algorithm based on successive convex approximation(SCA).For the latter,this thesis proposes a novel method to design the traversal order,which can flexibly determine the order according to the size of the data captured by the UAV at the cruise point.The proposed scheme comprehensively considers the dual effects of flight distance and communication time,and is compatible with both the traditional Travelling Salesman Problem(TSP)scheme and the Energy-Efficient Traveling Salesman Problem(EETSP)scheme,which greatly improve system performance.This thesis further investigates the time minimization problem for a cellularconnected UAV fixed-point patrol 3D scene.Different from the energy minimization problem,this problem expands the two-dimensional scenario to a three-dimensional one,the UAV can change altitude to improve the communication rate with the base station and reduce the data offloading time.In order to be more similar to the real environment and reflect the effectiveness of the proposed algorithm,no-fly zones are set in the simulation environment.Similarly,the time minimization problem can still be divided into the optimal transmission strategy between two consecutive cruise points and the traversal order design.For the problem of optimal transmission strategy between two consecutive cruise points,this thesis achieves it by minimizing the time and solving its dual problem,respectively.The non-convex problem is solved by introducing relaxation variables and successive convex approximation technique.The final simulation results show that the proposed solution outperforms other benchmark solutions,and the optimized task completion time is significantly reduced.