SDN-based approach for adaptive reconfiguration of routing in IoT for smart-buildings

Software Defined Networking (SDN) is a widely adopted approach for wireless network reconfiguration. However, its use in IoT context is still challenging. A fundamental issue is related to the expensive overhead of SDN signaling in an IoT architecture where the lowest level is composed of a multi-hops Wireless Sensor Network (WSN). This work aims to contribute on overcoming such issues. Precisely, we investigate the improvement of SDN efficiency in the case of dynamic reconfiguration of Low-Power and Lossy Networks (RPL) parameters. To conduct our study, we focus on mu SDN, a lightweight SDN architecture designed for IoT environments and developed under Contiki framework. Three key enhancements are introduced to mu SDN. First, a mechanism that regulates SDN signaling messages accordingly with the state of network topology. Second, a proactive path establishment to mitigate network access delay for scheduled traffic requests. Third, a dynamic configuration of RPL to adjust the energy consumption accordingly with two modes: a day mode where human activity in the smart building is intense, and a night/holiday mode where less traffic is generated by the sensors. Performances evaluation of our version, referred to as e-mu SDN (enhanced mu SDN), shows that our proposal significantly reduces the overhead of SDN signaling messages and the induced energy consumption while improving the Packet Delivery Ratio. We also show the ability of e-mu SDN to dynamically adjust RPL parameters following the day and night modes.