Security framework for supply chain IoT devices based on dynamic watermarking algorithm

• Main Author: Gu, Lei
• Other Authors: Chen Songlin
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Industrial engineering
• Online Access: https://hdl.handle.net/10356/172202
• Institution: Nanyang Technological University

The role of Internet of Things (IoT) devices in the supply chain is becoming increasingly significant, but at the same time, they face escalating security threats. Due to the characteristics of IoT device networks and the evolving attack methods, traditional cybersecurity measures offer limited protection for IoT device networks in the supply chain. Therefore, it is necessary to develop a dedicated security framework. This dissertation proposes a security framework that includes the dynamic watermarking system and the anomaly cause analysis system running on the gateways of the IoT device network. The former uses the Long Short-Term Memory (LSTM) method to achieve dynamic watermark encryption and decoding on signal packets, protecting the transferred information from being leaked. It also includes an anomaly detection module to determine the trustworthiness of received signal packets and their corresponding sending devices. The latter builds a game between the IoT device networks and attackers, determining the weak points and root causes of anomalies based on the optimal choices of both players, which helps supply chain managers allocate defense resources rationally and reduce the risk and losses of being attacked. To test the performance of the proposed security framework, we conducted a simulation experiment. The results indicate that the LSTM model training for the dynamic watermarking system performs excellently, with the detection module achieving the required performance metrics in different conditions. The anomaly cause analysis system can provide the system's root anomaly cause and weakest link based on historical data, thereby improving the security level of IoT devices in the simulated supply chain system.