Intrusion Detection System for Iot Devices Using CNN-LSTM Based Hybrid Framework

Abstract

The proliferation of Internet of Things (IoT) devices has significantly expanded the attack surface for cybercriminals, necessitating robust intrusion detection systems (IDS). Traditional machine learning-based IDS often struggle with adaptability, scalability, and detection of sophisticated attacks. This paper proposes a hybrid deep learning model combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks for IoT intrusion detection. The CNN extracts spatial traffic features, while the LSTM captures temporal dependencies. Extensive experiments were conducted on NSL-KDD, benchmarking the hybrid CNN-LSTM model against standalone CNN, LSTM methods. These findings highlight the feasibility of deploying hybrid deep learning IDS on resource-constrained IoT edge devices.

KEYWORDS: - IoT Security, Intrusion Detection System, Deep Learning, CNN-LSTM, Adversarial Robustness, Edge Computing