Providing context as a service using service-oriented mobile indie fog and opportunistic computing

Abstract

The increasing number of sensor-embedded mobile devices has motivated the research of mobile Sensing as a Service in which mobile devices can host Web servers to serve sensory data to the Internet of Things systems, urban crowd sensing systems and big data acquisition systems. Further, the improved processing power of modern mobile devices indicates the mobile devices are not only capable of serving sensory data but also capable of providing Context as a Service (CaaS) based on requesters’ own interpretation algorithms. In order to demonstrate mobile CaaS, this paper proposes a service-oriented mobile Indie Fog server architecture, which enables dynamic algorithm execution and also supports distributed CaaS processing among mobile devices. Moreover, in order to optimise the process distribution, the proposed framework also encompasses a resource-aware process assignment scheme known as MIRA. Finally, the authors have implemented and evaluated the proposed framework on a number of real devices. Accordingly, the evaluation results show that the MIRA scheme can improve the process assignment in the collaborative mobile CaaS environment.