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 ac-cordingly 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.