Trust based task offloading scheme in UAV-enhanced edge computing network

Abstract

Unmanned Aerial Vehicle (UAV) with a server which has powerful computing ability can fly over Internet of Thing (IoT) devices and conduct task offloading, giving rise to the so-called UAV-enhanced edge computing network. However, energy efficient and trustworthy UAV-enhanced edge computing is still a challenging issue. In this paper, a UAV-Trust based Task Offloading (UAV-TTO) scheme is proposed to offload tasks in an energy efficient and reliable way for IoT devices. The main innovations are as follows: (1) A Cluster based Task Offloading (CTO) approach is proposed in which tasks only need to route to any IoT devices within a cluster. The disadvantage of long flight distance and large energy consumption for UAV in traditional method can be overcame. Besides, the CTO approach avoids the far routing distance to the cluster head and the excessive data load. (2) A Traceback based Trust Evaluation (TTE) mechanism is proposed to evaluate the trust of devices. In this mechanism, the UAV can collect the task forwarding information provided by IoT devices when flying over the cluster. Then, we conduct a trust evaluation and reasoning mechanism according to the trust evidence collected to obtain more accurate evaluation results. IoT devices with high trustfulness will be selected to participate in task offloading, malicious devices will be excluded, so the efficiency of task offloading can be significantly improved. The UAV trajectory optimization strategy is also proposed to assist task offloading by considering trust and energy consumption. Extensive experimental results demonstrate that the proposed UAV-TTO scheme can reduce the total energy consumption for accomplishing the tasks effectively, evaluate the trust of IoT devices accurately and improve the success rate of task offloading.