A software-defined model for IoT clusters: Enabling applications on demand

Sensors and smart devices have become in-dispensable elements in the Internet of Things (IoT) systems. Collectively, they interconnect in vast number and provide massive amount of data about themselves and their physical environments. More importantly, numerous IoT services are envisaged to serve us in various capacities from personal health, smart homes to smart cities and critical infrastructures. Current IoT systems are mostly application-specific, rigidly configured and non-programmable and hence their applications are limited and difficult to reuse. The challenge is in handling a large number of sensors/IoT devices, many with limited capability, into intelligent, useful services on-demand. Software-defined concept of Software-Defined Networking (SDN) and software implementation of networked devices of Network Functions Virtualization (NFV) promise an effective solution to the challenge. This paper proposes a software-defined IoT (SD-IoT) model through leveraging these two emerging SDN and NFV technologies. In particular, the paper presents the design of a streamline SD-IoT controller, a lightweight and reconfigurable software-defined virtual sensor (SD-VSensor) to represent its underlying sensors, and a communication protocol (S-MANAGE) between them. The implementation results demonstrate the feasibility and efficiency of the proposed architecture.