Three Years of Crowdsourcing Network Traffic from Smart Homes
summary
An open-source tool called IoT Inspector collects three years of smart-home network metadata using ARP spoofing and crowdsourced labels, amassing traffic from 63,000+ devices across ~6,400 users. The team reports per-flow metadata only (no payload), a median run time of ~40 minutes, ~25% device labeling, messy label quality, missed DNS/SNI visibility that hides some endpoints, and work-in-progress features like blocking and Raspberry Pi deployment to increase duration and actionable user controls.
AI-written summary · about
chapters (5)
related talks
06mIoT Inspector: Crowdsourcing Labeled Network Traffic from Smart Home Devices at Scale
05mIoT Inspector: a platform for real-world smart home research
30mThe Hunt for Major League IoT-ICS Threats: A Deep Dive into IoT Threat Terrain
40mSecurity/Privacy in the IoT generation
49mLet's See What's Out There - Mapping the Wireless IOT
40mUncovering Covert Channels in Your IoT Networks
32mOpen Source IoT at Home
12mTransparent Microsegmentation in Smart Home IoT Networks