ROBUST AND SECURE IOT DEVICE AUTHENTICATION USING ON-CHIP PHYSICAL UNCLONABLE FUNCTIONS

Abstract

The Internet of Things' (IoT) explosive growth has highlighted the need for hardware-anchored, lightweight, and scalable authentication techniques to safeguard billions of linked devices. Conventional encryption methods are unsuitable in IoT environments with restricted resources because they frequently result in high computation overhead and energy consumption. In this paper, a secure IoT authentication system based on an Encoder-MUX Based Bistable Ring Physical Unclonable Function (EMBR PUF) challenge-response scheme is presented. The suggested authentication system integrates an architecture-aware challenge-response obfuscation method inside the EMBR-PUF structure to increase resistance against modeling and replay attacks. Through encoder–MUX interactions, the applied challenge dynamically regulates the internal propagation paths, creating hidden path-dependent behavior and making the challenge–response mapping more non-linear. Nonce-assisted challenge-response obfuscation is also used to improve session-level security with a lightweight approach appropriate for Internet of Things devices. The study concentrates on mutual authentication, session key formation, and replay attack resistance in IoT networks to broaden its usefulness. Using EMBR PUF-derived challenge-response pairs, the developed method offers a mechanism to create distinct device IDs and produce a session key without using resource-intensive cryptographic primitives. Thus, the system presents a three-phase exchange protocol between a device, a gateway and the PUF to achieve mutual authentication with low latency and minimal energy overhead on the gateway. Performance evaluation of the protocol implemented on an Artix-7 FPGA demonstrates near-optimal results, achieving 48.91% uniqueness, 97.22% reliability, and 50.63% uniformity, indicating strong device differentiation. Additionally, it exhibited low communication overhead, equating to 450 bits, and furthermore outperformed existing IoT authentication protocols (including DTLS, HIP, and MAP). By operating at the system level and utilizing the EMBR PUF, the implemented design provides a lightweight, hardware assisted and scalable scheme for IoT security for next generation ecosystems.