A context-aware IoT architecture through software-defined data plane

Current host-centric Internet infrastructure is inefficient in supporting data-oriented Internet-of-Things (IoT) applications, where contextual information of data packet is desirable for performing in-network forwarding and processing. In this paper, we prototype a context-aware IoT architecture that can forward and process IoT traffic in data plane based on contextual information exposed from both high application-layer and low sensor-layer, to fill the gap between IoT and IP network leveraging SDN and NFV through software-defined data plane. We implemented the proposed IoT architecture including programmable MVNO switch on many-core processor and IoT gateway with Intel Edison board and show the feasibility of our proposal with example use cases.