Survey on Intrusion Detection Systems based on Deep Learning

Abstract

Intrusion Detection Systems (IDSs) have a significant role in all networks and information systems in the world to earn the required security guarantee. IDS is one of the solutions used to reduce malicious attacks. As attackers always changing their techniques of attack and find alternative attack methods, IDS must also evolve in response by adopting more sophisticated methods of detection. The huge growth in the data and the significant advances in computer hardware technologies resulted in the new studies existence in the deep learning field, including intrusion detection. Deep learning is sub-field of Machine Learning (ML) methods that are based on learning data representations. In this paper, a detailed survey of various deep learning methods applied in IDSs is given first. Then, a deep learning classification scheme is presented and the main works that have been reported in the deep learning works is summarized. Utilizing this approach, we have provided a taxonomy survey on the available deep architectures and algorithms in these works and classify those algorithms to three classes, which are: discriminative, hybrid and generative. After that, chosen deep learning applications are reviewed in a wide range of fields of intrusion detection. Finally, popular types of datasets and frameworks are discussed.