Publications
ExypnoSteganos - A Smarter Approach to Steganography
April 2021
Journal of Intelligent & Fuzzy Systems - IOS Press
With the ever-rising threat to security, multiple industries are always in search of safer communication techniques both in rest and transit. Multiple security institutions agree that any systems security can be modeled around three major concepts: Confidentiality, Availability, and Integrity. We try to reduce the holes in these concepts by developing a Deep Learning based Steganography technique. In our study, we have seen, data compression has to be at the heart of any sound steganography system. In this paper, we have shown that it is possible to compress and encode data efficiently to solve critical problems of steganography. The deep learning technique, which comprises an autoencoder with Convolutional Neural Network as its building block, not only compresses the secret file but also learns how to hide the compressed data in the cover file efficiently. The proposed techniques can encode secret files of the same size as of cover, or in some sporadic cases, even larger files can be encoded. We have also shown that the same model architecture can theoretically be applied to any file type. Finally, we show that our proposed technique surreptitiously evades all popular steganalysis techniques.
Dec 2019
Security in Computing and Communications - Springer - Germany
With the wide adoption of the internet and its applications in recent years, many antagonists have been exploiting information exchange for malicious activities. Intrusion detection and prevention systems are widely researched areas, rightly so being an integral part of network security. Adoption of IDSs and IPSs in networks have shown significant results while expanding research from software solutions to hardware-based solutions, promoting such defensive techniques even further. As with all recent computing trends, Machine Learning and Deep Learning techniques have become extremely prevalent in intrusion detection and prediction systems. However, intrusion prediction is still in its infancy. There have been attempts but none projecting any significant improvement over the current systems. Traditional systems alert the user after an intrusion has occurred, steps can be taken to stop further expansion of the intrusion, but in most cases, it is too late. Hence catering to this issue, this paper proposes system call prediction using a Recurrent Neural Network (RNNs) and Variational Autoencoding modelling techniques to predict sequences of system calls of a modern computer system. The proposed model makes use of ADFA intrusion dataset to learn long term sequences of system-call executed during an attack on a Linux based web server. The model can to effectively predict and classify sequences of system-calls most likely to occur during a known or unknown (zero-day) attacks.
April 2019
International Conference on Recent Trends on Electronics, Information, Communication & Technology - IEEE
With the wide adoption of the internet and its applications in recent years, many antagonists have been exploiting information exchange for malicious activities. Intrusion detection and prevention systems are widely researched areas, rightly so being an integral part of network security. Adoption of IDSs and IPSs in networks have shown significant results while expanding research from software solutions to hardware-based solutions, promoting such defensive techniques even further. As with all recent computing trends, Machine Learning and Deep Learning techniques have become extremely prevalent in intrusion detection and prediction systems. However, intrusion prediction is still in its infancy. There have been attempts but none projecting any significant improvement over the current systems. Traditional systems alert the user after an intrusion has occurred, steps can be taken to stop further expansion of the intrusion, but in most cases, it is too late. Hence catering to this issue, this paper proposes system call prediction using a Recurrent Neural Network (RNNs) and Variational Autoencoding modelling techniques to predict sequences of system calls of a modern computer system. The proposed model makes use of ADFA intrusion dataset to learn long term sequences of system-call executed during an attack on a Linux based web server. The model can to effectively predict and classify sequences of system-calls most likely to occur during a known or unknown (zero-day) attacks.