Analysis of Hybrid Deep Learning Algorithms for Distributed Denial of Service Attack Detection in the Internet of Things

Abstract

The Internet of Things (IoT) plays a pivotal role in shaping smart city environments by enabling the interconnectivity of various devices over the internet backbone. However, ensuring the security of IoT systems is of paramount importance, given their vulnerability to diverse cyber threats, including Distributed Denial of Service (DDoS) attacks. This survey focuses on the classification and detection of DDoS attacks within IoT ecosystems, employing a hybrid approach incorporating Deep Learning (DL) algorithms. The analysis reveals that hybrid classifiers outperform other methods, showcasing superior performance across essential performance metrics. These optimized hybrid classifiers, primarily evaluated on accuracy, demonstrate heightened efficiency in detecting attacks within IoT environments. Notably, the most frequently employed classifiers for attack detection are the Hybrid Convolutional Neural Network (Hybrid CNN) and Long Short-Term Memory (LSTM) models. In summary, our survey examines the landscape of hybrid DL algorithms used for IoT attack detection, emphasizing performance metrics, publication timelines, methodology diversity, and notable achievements in this critical field of IoT security.