| With the rapid development of Internet of things(Io T)and 5G communication technology,the intelligent devices connected to the Internet has increased explosively.Given the application service latency and network communication cost constraints,these devices need to store and process data with the help of edge computing platforms to mine information value from the massive data to serve various Io T applications.However,for computation-intensive tasks,the devices cannot process them effectively in real time with weak energy and computing power.Therefore,offloading device tasks to mobile edge computing platform nodes(e.g.,UVA,vehicles)for computation is a feasible and effective way in Io T applications.However,in large-scale networks,the excessive mobile overhead and communication overhead of mobile edge computing platform nodes traversing devices to collect tasks shortens their duration of operation,which is not conducive to maintaining the desired network lifetime and guaranteeing the timeliness of tasks.Therefore,minimizing the system overhead while satisfying the task offloading delay constraint is an important scheme goal to be achieved.Secondly,devices in the Io T are deployed by multiple parties,and trustworthiness cannot be fully guaranteed.Therefore,malicious device nodes may participate in network data transmission,where discarding or tampering with the data to be forwarded,thus reducing the offloading utility.Therefore,identifying and screening out malicious devices in the offloading process is another key issue.Above all,this thesis proposes a Trust and Effective Task Offloading Scheme(TETO)for Io T applications.TETO not only improves the system energy utilization and task timeliness,but also effectively detects malicious nodes in the network and protects data security.The main research work of this thesis is as follows:(1)Aiming at the task offloading problem in Io T,TETO is proposed to reduce system energy consumption and task processing delay,and an adaptive network clustering algorithm was designed,which comprehensively considers device residual energy and task data amount.Based on the real-time situation of the network,the fuzzy logic is applied to realize the adaptive selection of cluster head nodes.The proposed TETO scheme simplifies the UAV flight trajectory and reduces the UAV overhead through network clustering and task data aggregation,and also selects the devices with more remaining energy and larger task data volume as cluster head nodes,which helps to balance the network load and reduce the amount of network data transmission.(2)To address the device trustworthiness problem,this thesis proposed an active chain trust inference mechanism.The mechanism uses UAVs to collect task data forwarding information on the routing path in passing while collecting tasks,and identifies and monitors suspicious nodes based on whether the devices submit contradictory information.Secondly,we propose a trust evolution scheme to update the trust of the device in real time,and take the trust as a new factor in the selection of cluster head,so as to avoid malicious nodes becoming cluster head nodes and preventing the task data of the whole cluster from being jeopardized.(3)To further reduce the overhead,we designed a heuristic UAV path planning algorithm to optimize the flight trajectory on the premise of greatly reducing the traversing node number.The algorithm continues to optimize the flight trajectory with a significant reduction in the number of traversing nodes.In order to collect sufficient task forwarding information,the algorithm offsets the trajectory with lower overhead increments based on the shortest path,thus acquiring relevant data for trust inference after a complete routing path,which allows the trust evolution of the device to converge as soon as possible.Theoretical analysis and experimental results show that the use of TETO in Io T applications can guarantee significant performance improvements in terms of energy consumption,latency and security for task offloading. |
PDF Full Text Request