Abstract
Mobile Edge Computing can provide application-centric Internet of Things (IoT) services at the brim of the user's network. This reduces the latency and increases the availability of the services for the users. However, security measures are limited to the density of adversaries detected in the network's edge. Impeding adversaries increase the threats for privacy and service sharing through eavesdropping or jamming. By considering the adversaries' threat in the network's edge, this paper introduces privacy-controlled offloading scheme for improving the data sharing security between the edge-IoT devices. In this scheme, harmonized trust validation using recursive decision making is performed, followed by the data offloading process. The trusted non-virtual edge devices are identified based on the previous successful sharing instances. Post to detect non-virtual trusted edge devices, and the offloading is performed in privacy preserved sessions exploiting volatile shared key authentication. Verifying the edge devices' trust by differentiating the virtual and real is facilitated using deep belief networks. The proposed scheme is reliable in preventing service loss (5.26%), reducing false alarm (0.2%), and improving the network's availability of trusted devices (96.2%).