April 5, 2024
Elevating Industrial IoT Security with LSTM-Based Wi-Fi Jamming Detection
In the rapidly expanding IoT landscape, Wi-Fi serves as the communication technology for countless applications, demanding robust security measures to counteract jamming threats. The Long Short-Term Memory (LSTM) based jamming detection and forecasting model developed and implemented by MarUn, is meticulously designed to protect Wi-Fi-based IoT systems by analyzing transport and application layer parameters. In this research, we initially explored the impact of jamming attacks on the upper layers of the network protocol stack within IoT communications. Subsequently, these identified parameters served as the foundation for training our model. Our model offers real-time jamming detection with exceptional accuracy, preserving the integrity and reliability of IoT communications. The model is trained to detect various kinds of jamming attacks to safeguard against both conventional and sophisticated jamming techniques.
Through rigorous testing in a simulated industrial IoT environment, the model demonstrated a remarkable detection accuracy of 99.5% and precision of 99.4%. These results underscore the model's efficacy in maintaining uninterrupted IoT operations, even in the face of targeted jamming disruptions.
You can access the paper using following link.
