SDN-Enabled IoT: Ensuring Reliability in IoT Networks Through Software Defined Networks

Ensuring reliability for IoT networks is very crucial for the use cases like autonomous self-driving car, tactile internet, healthcare devices, etc., which requires continued communication to facilitate un-interrupted services. Software defined networks (SDN) facilitates to program the network and enables efficient control over the complicated network infrastructure like IoT. For a continued and effective implementation of SDN in IoT networks, it must solve the network reliability challenges to provision the low-latency and ultra-reliable transmission scenarios even in the case of failures in the network. In comparison with the path-based recovery, the local rerouting is a preferred solution for rapid failure recovery. For achieving the rapid local recovery, backup paths must be pre-configured for every flow on the link, which results in memory consumption of the switch for maintaining flow rules of the backup paths. Also, the efforts required for rerouting of every flow can delay the failure recovery. The book chapter will focus on the issues associated with failure management in software defined IoT networks and proposes forwarding table configuration in network, which can autonomously recover an OpenFlow-based IoT network from a link or a node failure. We firstly present Local Immediate (LIm) and Immediate Controller Dependent (ICoD) failure management approaches to overcome the shortcomings of link failure management approaches in SDN. Our proposed approaches conserve the memory of switches by reducing the backup path rules by aggregating the flow rule on the common network component using VLAN-enabled flow labelling. The proposed approaches are expected to accomplish recovery in the range of 2–20 ms and will fulfill the stringent 50 ms recovery condition of Carrier Grade Networks (CGNs). Next, we extend our solution on single link recovery and present forwarding table configuration for the network to accomplish the switch recovery. To validate our proposed approaches for the link and switch recovery, we evaluated the performance in following points; (i) Number of flow entries can be saved to enable the protection against failures, and (ii) how quickly the recovery can be accomplished.