Characterization of Google Home Traffic in a Home Network

Abstract

IoT devices are currently getting into our everyday life. However, most of the people are not aware of the information that is being shared by these devices on the Internet, or how do they impact their home network. We used a software-defined network within a real home environment to collect, analyze and characterize the traffic generated by Google Home. We used the Zodiac FX switch, a low-cost OpenFlow-compliant switch. After the data collection and analysis, we characterize Google Home’s behavior to provide insight on traffic behaviour, bandwidth consumption and commonly used protocols. Our results confirm what recent studies have indicated: a network provider can easily identify the presence of a Google Home in the network and can identify whether or not the device is being used. Thus, we propose how to leverage a low-cost SDN home network to reduce the amount of private information that is leaked to the service provider.