JIITA, vol.7 no.4, p.874-888 2023, DOI: 10.22664/ISITA.2023.7.4.872
Tatiparthy Ramesh reddy, D. Usha
Abstract: Due to its successful use in many traditional artificial intelligence (AI) problems as compared to standard ML algorithms, Deep Learning (DL), a novel form of machine learning, is attracting a lot of research interest (CMLAs). For a variety of applications in the field of cyber security, DL architectures have recently been creatively developed. As researchers explore various cutting-edge DL models and prototypes that may be customised to fit particular cyber security applications, the literature on DL architectures and their modifications is expanding. A thorough review of these research studies is however lacking in the literature. Numerous survey based studies lack a futuristic evaluation and instead concentrate on certain DL designs and specific harmful attack types within a constrained cyber security problem scenario from the past. With regard to next-generation cyber security scenarios involving intelligent automation, the Internet of Things (IoT), Big Data (BD), Blockchain, cloud, and edge technologies, this article intends to provide a comprehensive and well-rounded survey of the past, present, and future DL architectures. This work compares and analyses the contributions and difficulties from many recent research papers to give a tutorial-style thorough analysis of the state-of-the-art DL architectures for various applications in cyber security. First and foremost, the survey is distinctive in that it reports the use of DL architectures for a wide range of cybercrime detection techniques, including intrusion detection, malware and botnet detection, spam and phishing detection, network traffic analysis, binary analysis, insider threat detection, CAPTCHA analysis, and steganography. Second, the survey discusses important DL designs in areas of cyber security such encryption, cloud security, biometric security, IoT, and edge computing. Thirdly, the demand for DL-based research for the next-generation cyber security applications in cyber physical systems (CPS) that leverage BD analytics, natural language processing (NLP), signal and image processing, and blockchain technology for smart cities and Industry 4.0 of the future is discussed. Finally, a critical analysis of current issues and the new DL design that has been presented advances the field of study.
Keywords: Cyber Security, Machine Learning, Neural Networks, Deep Learning, Communication Networks, Cloud and Edge Computing.
Fullpaper: