Exploiting state information to support QoS in Software-Defined WSNs

Software-Defined Networking (SDN) is a promising paradigm initially proposed for the wired networks which has the potential to change the way in which networks are designed and managed. Recently, its use has been extended to wireless infrastructure-less networks such as ad-hoc and wireless sensor networks. In this paper, we take inspiration from our previous work where a scheme that extends the SDN paradigm to WSNs - called SDN-WISE - was proposed. Differently from OpenFlow, SDN-WISE is stateful; accordingly in this paper we investigate how state can be used to support different levels of QoS in WSNs. In fact, state can be used to represent the level of congestion of a node and rules can be given to the nodes that set different drop probabilities for different flows depending on the level of congestion. By reporting their state, the sensor nodes can inform the Controller about their level of congestion; thus, the Controller can select and set alternative routes for traffic flows in order to mitigate congestion. Numerical results assess the effectiveness of the proposed scheme.