From Evaluating to Enabling SDN for the Internet of Things

The increase of the number of smart connected things, Internet applications and services has raised new challenges in the networking issue. Dynamic management, control and programmability are the key challenges of new networking architectures such as the Internet of Things (IoT) architecture. Traditional networking methods based on static configuration and management, caused by vertical integration, can not support new communication technologies requirements. Software Defined Networking (SDN) came as a solution that separates the control plane from the data plane. It enables the network to adapt its behavior dynamically according to the traffic type for example. The SDN controller is the brain and the most important element in the SDN-based architecture. It has a centralized and a complete knowledge of the network and the flexibility to tune the components and especially in the IoT environment. In this paper, firstly, we analyze and compare the performances of the SDN-based with the traditional networking architecture. Secondly, the most used SDN controllers (ONOS, OpenDayLight, Ryu and Pox) are compared and Ryu is selected according to its performances regardless of the three chosen architecture topologies (single, linear and tree). Finally, we apply SDN with the selected controller to an IoT architecture in order to evaluate its feasibility and its performances using IoT traffic flow transmitted through MQTT protocol.