Resilient Software Integrity for IoT Networks

Abstract

Internet-of-Things (IoT) devices have become increasingly pervasive across modern society, often performing safety-critical functions. Network attestation – verifying the software state of all devices in a network – is a promising mitigation approach. However, current network attestation schemes suffer from lengthy Time-Of-Check-Time-Of-Use (TOCTOU) vulnerability windows, high latency, and interference from compromised devices. In this paper, we present TRAIN, a TOCTOU resilient attestation technique that synchronizes network-wide attestation, enables constant-time per-device reports, and remains resilient even when multiple devices are compromised.