A SDN-Based Approach for Conflicting Energy Profiles on Partially Sustainable IoT Networks

Equipping millions of Internet of Things (IoT) devices with rechargeable batteries and energy ambient harvesters introduce a significant maintenance cost. Thus, it is realistic to expect that the IoT networks will become greener gradually, over the next years, which means that current IoT networks are a mixed deployment with sustainable and traditional (i.e. non-sustainable) devices. This gradual transformation results in partially sustainable networks, where the traditional nodes have an energy profile to minimize energy spent, while the sustainable nodes must seek neutral or self-sustained energy consumption. Regularly, these energy profiles might be in conflict. This work addresses this challenging scenario proposing a solution that deals with the conflicting energy profiles for partially sustainable IoT networks. To achieve this, the proposed approach is based on Software-Defined Networks, where a controller is responsible for defining forward rules and data aggregation levels for the network traffic. The solution also applies local two-stage stochastic programming to set the data aggregation timer and minimize the energy cost per payload unit. The obtained results show the performance of the proposed solution for abundant and scarce energy resources, measuring the energy consumption and residual energy as indicators of node sustainability.