Internet of Things (IoT) networked systems are experiencing exponential growth, with billions of devices like cameras, sensors, and actuators deployed globally. These devices are used in critical infrastructures across sectors such as health, manufacturing and energy. However, relatively low-cost IoT devices sourced from various manufacturers are not equipped with adequate embedded security measures and continue to face sophisticated cyber threats.

However, operators of such networks may not even be fully aware of their IoT assets, let alone whether each IoT device is functioning properly and safe from cyber-attacks. We address these challenges by developing a robust framework for IoT device classification and attack surface management, leveraging characteristics of behaviors exhibited at the network level.

We instrument a smart environment with tens of different types of IoT devices, including cameras, lights, plugs, motion sensors, appliances, and health monitors. For years, we have been collecting traffic traces from this infrastructure, and we are releasing a subset of these traces as open data for the community’s use. Additionally, we present insights into the underlying network traffic characteristics using statistical attributes such as activity cycles, port numbers, signaling patterns, and cipher suites.