SHIOT: A novel SDN-based framework for the heterogeneous internet of things

Abstract

A new measurable and quantifiable world is created by the Internet of Things. However, the variety of IoT components, i.e., devices, access technologies and applications, which are deployed on the same core infrastructure with a common network policy have led to an unexpected issue of heterogeneity. Such issue directly dismisses the interoperability in the IoT, and hence, significantly decreasing the QoS of a given IoT service. In this paper, we develop a SDN-based framework, called SHIOT to address the above challenge. SHIOT relies on the ontology for examining the end-user requests and applies a SDN controller to classify flow scheduling over the task level. We also utilize the Lagrange relaxation theory to optimize the routing mechanism. Extensive experiments demonstrate that SHIOT is able to support stressed networks and offers a significant advantage over the traditional framework that is integrated without SDN.