Research On Multi-Layer Network Computing Offload Assisted By Unmanned Aerial Vehicle Group

With the continuous development of IoT technology,mobile computing is gradually undergoing a paradigm shift from centralized cloud computing to Mobile Edge Computing(MEC).The main feature of MEC is to transfer computing,storage,etc.to the edge of the network(such as access points and base stations)to meet the needs of computing-intensive and delay-sensitive applications on mobile user equipment with limited resources.Ground MEC systems are susceptible to interference from terrain obstacles and other external factors,resulting in unstable communication links.For this reason,Unmanned Aerial Vehicle(UAV)assisted MEC systems have gradually become one of the research hotspots in the field of edge computing.Due to the good characteristics of UAVs such as high flexibility,convenient deployment,and strong computing power,it is considered to have great development prospects in assisting ground wireless communications.Utilizing UAV-assisted MEC systems can not only effectively utilize communication resources,improve the performance of wireless communication systems,but also reduce the energy consumption of user equipment.However,factors such as difficult UAV path planning and complex pre-cache processing of request content limit its wide application in the field of edge computing.In addition,user equipment and drones are very sensitive to their own energy consumption,so optimizing the flight path of drones and content cache placement updates,and formulating efficient computing offload strategies are crucial to reducing the overall energy consumption of user equipment and drone servers.It is very important,and it is a pain point that needs to be solved urgently in the field of edge computing.This paper will conduct research on computing offloading in the three-layer computing environment of "terminal layer-edge server layercloud server layer",in which the UAV group acts as the edge server layer to provide computing offloading services.The main research contents are as follows:(1)Research on the computing offloading strategy based on the dynamic collaboration of UAV groups under the multi-layer networkDynamic collaboration,that is,the dynamic planning and mutual cooperation of the overall path of the UAV group.This paper proposes a method that combines edge computing technology and UAV group technology,and uses an alternate optimization algorithm to obtain an efficient computing offload strategy under the optimized UAV group flight path.This strategy can not only deal with the computing content offloaded by users,but also allow the UAV group to dynamically cooperate to achieve the best load balancing and task distribution,so as to reduce the overall energy consumption of user equipment and servers and prolong the overall service time.The simulation results indicate that compared to other computation offloading strategies,this computation offloading strategy can effectively reduce the total energy consumption of user devices and drone fleets by 9%to 23%.(2)Research on the calculation offloading strategy based on content cache placement update under multi-layer networkThis paper proposes a computing offloading strategy based on multiuser and multi-server structure under multi-content cache placement update.Combining edge computing technology and caching technology,the content caching of the edge network and the coordination of the core network are realized.Pass information between servers through the cache queue to speed up the processing of offloaded content.For different request content information,an optimized cache placement update method is used at each network layer to meet the needs of real-time update applications.Combining the greedy strategy with the alternate optimization algorithm,an efficient computing offloading strategy under content cache placement and updating is obtained,which reduces the overall energy consumption of user equipment and servers.The simulation results show that the proposed computation offloading strategy under the updated multi-content caching placement is able to achieve 12%to 25%lower energy consumption compared to existing offloading strategies in various scenarios.