Detecting Privacy Leakage of Smart Home Devices through Traffic Analysis

Abstract

Under the management of the Internet of Things platform, smart home devices can be operated remotely by users and greatly facilitate people’s life. Currently, smart home devices have been widely accepted by consumers, and the number of smart home devices is rising rapidly. The increase of smart home devices introduces various security hazards to users. Smart home devices are vulnerable to side-channel attacks based on network traffic. The event of smart home devices can be identified by network surveillants. Given this situation, we designed a set of standardized workflows for traffic capturing, fingerprint feature extraction, and fingerprint event detection. Based on such workflow, we present IoTEvent, a semiautomatic tool to detect vulnerable smart home devices, which is not limited to specific types of communication protocols. IoTEvent first collects device traffic by simulating touch events for App. Then, it pairs the packet sequences with events and generates a signature file. We also test the usability and performance of IoTEvent on five cloud platforms of smart home devices. Finally, we discuss the reasons for privacy leakage of smart home devices and security countermeasures.