Automation of IoT Network Traffic Analysis and Anomaly Detection using Deep Packet Inspection with Hybrid DenseNet121-BiLSTM
Geo Francis E, S. Sheeja, E. F. Antony John, Jismy Joseph · 2025
Network security is one of the most crucial problems facing in Internet of Things. An essential component of information systems' long-term viability and regular functioning is network intrusion detection. One potential technique that aids in the identification and defense against network threats is deep packet inspection (DPI). A novel DPI method utilizing a blend of LSTM, GRU, and Recurrent Neural Networks is presented in the paper. The proposed model enhances the accuracy and reliability of detecting outcomes for the threats included in network traffic data. The proposed system detects and identify particular threats, such as malware exploitation, DoS attacks, and unauthorized access attempts, allowing linked devices and services with an accuracy of 99.99%. The system, therefore, can be applied in real-time monitoring systems, intrusion detection and prevention systems, and forensic investigations. Through a series of quantitative trials and comparisons with the most recent formulations, the effectiveness of this approach is assessed.