Research On UAV-Assisted Task Offloading And Energy Consumption Optimization

With the advent of the 5G era,the demand for Quality of Service(Qo S)of Internet of Things Devices(IDs)is increasing gradually.Mobile Edge Computing(MEC)is proposed to provide stable and low-latency network services for IDs.However,in hot spots,remote areas and emergency areas,edge servers can’t meet the communication needs of IDs.Nowadays,due to the high flexible deployment ability,and low cost of Unmanned Aerial Vehicles(UAVs),which could be considered as the assistant of MEC to collect and process the computation tasks of IDs that can’t be connected to edge servers.This thesis focuses on the UAV-assisted MEC task offloading and data collection scheme.This thesis studies the UAV-assisted task offloading scheme.Firstly,the UAV with computing resources is regarded as the Edge Nodes(ENs),and IDs with offloading tasks are classified as the Request-IDs(R-IDs)and the Free-IDs(F-IDs)with idle computing resources,which can provide offloading services for the former.In addition,Device-toDevice(D2D)and Ground to Air(G2A)links are considered to offload tasks parallelly.Under the premise of satisfying the Qo S requirements of R-IDs,the UAV trajectory,transmission power and task offloading ratio are jointly optimized to minimize the total network overhead.Since the original optimization problem is a nonconvex optimization problem,which is difficult to solve.It is divided into two subproblems: Resource Allocation optimization and UAV Trajectory Optimization,which are solved alternately by resource allocation(RA)algorithm with fixed UAV trajectory,and UAV trajectory optimization(UTO)algorithm until convergence.Finally,based on the simulation results,the effectiveness of the proposed algorithm is verified.Further,this thesis studies the scheme of UAV swarm assisted data collection.First of all,consider remote areas,where ground Sensor Nodes(SNs)can’t transmit data to the control center,and UAV swarm collects their data and forwards it to the control center.Secondly,the energy consumption optimization problem of UAV swarm assisted data collection is constructed,and the total energy consumption of the network is minimized by jointly optimizing the number of UAVs,hovering position,flight trajectory and the association between UAV and SNs.Because the original problem is nonconvex,to reduce the computation complexity,the optimization problem is transformed into sub-problems such as UAV deployment optimization,UAV trajectory optimization and UAV-SNs association optimization.Thirdly,the UAV Deployment(UD)algorithm is proposed to optimize the number and hovering position of UAV swarm.Moreover,considering the obstacle avoidance problem of UAV swarm,based on Dubins theory,we abstract the obstacle as a Dubins circle,and propose an improved shuffled Shuffled Frog Leading Algorithm(ISFLA)based on genetic algorithm to find the shortest obstacle avoidance trajectory of UAV swarm.Finally,a Multi-UAV Association(MUA)algorithm is introduced to solve the association between UAV swarm and SNs.The simulation results verify that feasibility of the proposed algorithms.