Arpotcam: augmented reality-driven honeypot for enhancing security in IoT surveillance systems

Arpotcam: augmented reality-driven honeypot for enhancing security in IoT surveillance systems

연구 분야: Infrastructure

학회: The Visual Computer

The proliferation of Internet of Things (IoT) devices, especially surveillance cameras, has transformed surveillance and monitoring across various domains. Moreover, augmented reality (AR) has guided immersive digital experiences. Honeypots have become vital tools in cyber security for detecting and analysing threats. This paper introduces ARPotCam, an innovative AR-based honeypot specifically designed for surveillance cameras. ARPotCam aims to convincingly emulate the behaviour of surveillance cameras and accurately respond to camera control commands, thereby creating a more deceptive and interactive environment for potential attackers. By integrating AR with cybersecurity for the IoT, ARPotCam enhances the realism of honeypots, making them more effective in detecting and analysing cybersecurity threats. It innovatively maps 360-degree video streams based on attacker-initiated camera commands, enhancing deception. Automated object prediction using reinforcement learning and real-world evaluations showcases the system’s scalability and effectiveness. We evaluated ARPotCam’s efficiency against reconnaissance and honeypot detection tools, shedding light on its covert capabilities. Through extensive simulations and experiments, we have gained valuable insights into the performance of our AR-enhanced surveillance camera honeypot. Evaluation results demonstrate that ARPotCam achieves a 72.5% deception rate, outperforming traditional honeypots by 20%. This system offers a scalable and adaptable approach to enhancing IoT security, particularly in surveillance systems.