Scaling the Internet to the Things: Combining Named Data Networking, Software-Defined Networking, and AI to Support the Next 50 Billon IoT Devices

The Internet of Things (IoT) has been growing at a great pace. It is enabling a wide range of applications that permeate many facets of our life. Examples include smart homes, healthcare, air and water quality monitoring, traffic control, and smart farming. Given the huge number of IoT devices that continue to be connected to the Internet, two questions arise: (i) how can the Internet be scaled to the size of the IoT, its increasingly open nature, and amount of traffic generated by its billions of devices? and (ii) how can this scalability be achieved while providing an adequate support to IoT devices given their typical characteristics, e.g., being battery powered and endowed with a moderate computing power? In this paper, we focus on answering the two previous questions. Specifically, we propose a novel architecture and query processing framework for the IoT where we combine named data networking (NDN), softwaredefined networking (SDN) and artificial intelligence (AI).