Analyzing energy consumption of smart home devices through network monitoring

Details

Supervisors

Barbette, Tom ; Sadre, Ramin

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] en sciences informatiques, à finalité spécialisée

Abstract

The number of smart homes and IoT devices is rapidly growing. In the meantime, a challenge for homeowners is to monitor the energy consumption of their appliances to reduce their consumption. While many smart devices don’t have built-in energy monitoring capabilities, they are connected to the local network of the house using Wi-Fi for example, and exchange information. These devices can be activated remotely, and when an activation event occurs it is observable in the network traffic. In this thesis, we develop an application that monitors the network traffic of these devices and identifies events in order to predict their energy consumption. We adapt a state-of-the-art approach to identify the state of each device (for example, detect if a smart light bulb is on or off). By combining the activities with the total energy consumption of the house, we create consumption profiles to predict and analyze their energy consumption. The results are satisfying for simple devices, even though there is a need for improvement due to limitations of the state-of-the-art approach that we used.