Hybrid Deep Learning Models for Intrusion Detection in IoT Networks Using CNN-LSTM Architectures

Abstract

The rapid expansion of the Internet of Things (IoT) has brought about unprecedented convenience and interconnectivity in sectors ranging from healthcare and industry to smart homes and cities. However, this expansion also introduces significant cybersecurity risks, making intrusion detection systems (IDS) an essential component of IoT infrastructure. Traditional IDS techniques often fall short in handling the dynamic and high-volume nature of IoT data. This paper explores a hybrid deep learning approach combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks for real-time intrusion detection in IoT networks. CNNs are utilized for extracting spatial features from packet-level data, while LSTMs are employed to capture temporal dependencies in data sequences. This hybrid model significantly improves the accuracy and robustness of anomaly detection compared to single-model architectures. The paper discusses the dataset used, preprocessing techniques, model design, evaluation metrics, experimental setup, and the comparative analysis of results. The study demonstrates the superiority of the CNN-LSTM model in identifying complex intrusion patterns, thereby enhancing real-time security in IoT networks.

Keywords: Deep learning, IoT, intrusion detection, cybersecurity, LSTM, CNN