Strengthening Industrial IoT Security: Device Fingerprinting and ML-Based Node Authentication at Gateway

Abstract

This paper aims to enhance the security of data transmission in small-scale industrial IoT systems using a novel approach combining static and dynamic device fingerprinting, sensor data fusion, and machine learning (ML) techniques. The objectives include creating static device fingerprints using ESP32 nodes, generating dynamic fingerprints based on sensor data variations, merging both fingerprints to create a unique fingerprint hash, and developing an ML model for node identification and data integrity verification. In the context of industrial IoT, where sensitive data is exchanged between devices and central monitoring systems, ensuring robust data transmission security is paramount. The increasing prevalence of cyber threats and potential vulnerabilities in industrial networks necessitates innovative security measures. This project addresses this critical need by leveraging dynamic fingerprinting and ML-based node authentication to establish secure and reliable data communication channels. The significance of this work lies in reducing the need for continuous authentication by leveraging dynamic fingerprinting and ML-based node authentication, thus enhancing overall system security. The key findings demonstrate the effectiveness of the proposed approach in securely transmitting data without compromising system integrity, thereby mitigating potential threats from unauthorized access and data tampering. This research contributes to the advancement of IoT security methodologies in industrial IoT environments, ensuring secure and reliable data communication, and strengthening defenses against evolving cyber threats.