Aura: An incentive-driven ad-hoc IoT cloud framework for proximal mobile computation offloading

Abstract

The rapid growth of mobile applications requires enhanced computational resources in order to ensure better performance, security, and usability. In recent years, the proliferation of the Internet-of-Things (IoT) devices has caused a paradigm shift in computing and communication. IoT devices are making our physical environment and infrastructures smarter, bringing pervasive computing to the mainstream. Given numerous predictions that we will have billions of such devices deployed in the next five years, we have the opportunity to utilize such IoT devices in converting our physical environment into interactive, smart, and intelligent computing infrastructures. In this paper, we present Aura – a highly localized IoT based cloud computing model. Aura allows mobile clients to create ad hoc and flexible clouds using the IoT and other computing devices in the nearby physical environment. Aura provides localized computational capability from untapped computing resources using a task-offloading model for mobile devices. Computations done in Aura are highly flexible, giving clients full control to start, stop, migrate, and restart computations in localized IoT devices as the mobile users move between different physical locations. As an example application of Aura, we have ported a lightweight version of MapReduce to run on IoT devices powered by Contiki OS. The prototype application was utilized to conduct various experimental measurements to evaluate different performance metrics of the proposed system. The paper presents a detailed comparative analysis of Aura with traditional clouds and applications running natively on mobile phones to assert the benefits and feasibility of the model.