ADVERSARIAL DEEP LEARNING FOR MALWARE ANALYSIS: DECEPTION AND COUNTERMEASURES

Project Summary: 

Malware detection is still one of the major problems in computer security. To fight against the increasing number of malware, their variability and sophistication, machine learning-based solutions have emerged as solution and it is being increasingly adopted by companies and institutions. In particular, deep learning have started to show [1,2] their impressive performance  on the area of malware analysis. However, while deep neural networks can provide state-of-art results on malware classification, they also vulnerable to adversarial examples [5] that can be created by slightly but cleverly manipulating the programs and binary files [3,4].

In this project, we propose the use of adversarial examples to, in a first instance, validate, and then, improve, the performance of conventional and machine learning-based malware detection systems. Specifically, we aim to generate new adversarial deep learning architectures and Generative adversarial networks (GAN) that could be used for attacking and defending security systems

[1] E. Raff, J. Barker, J. Sylvester, R. Brandon, B. Catanzaro, and C. Nicholas. Malware detection by eating a whole exe. arXiv preprint arXiv:1710.09435, 2017.

[2] McLaughlin, N., Martinez del Rincon, J., Kang, B., Yerima, S., Miller, P., Sezer, et al., G. Deep Android Malware Detection , ACM Conference on Data and Applications Security and Privacy (CODASPY) 2017

[3]  B.Kolosnjaji , A. Demontis, B.Biggio, et al. Adversarial Malware Binaries: Evading Deep Learning for Malware Detection in Executables, arXiv preprint arXiv: 1803.04173v1, 2018.

[4] A. Al-Dujaili, A. Huang, E. Hemberg, U. O’Reilly, Adversarial Deep Learning for Robust Detection of Binary Encoded Malware,  arXiv preprint arXiv: 801.02950v3, 2018.

[5] N. Carlini, D. Wagner, Towards Evaluating the Robustness of Neural Networks, ,  arXiv preprint arXiv: 1608.04644v2, 2017.

Objectives:

  • To investigate the effect of adversarial examples as obfuscation technique for malware intrusion and its effect on current anti-malware detection  
  • To investigate the use of GAN networks for both attacking and defending machine learning based  malware solutions.
  • To develop new adversarial deep learning architectures for malware analysis
  • To evaluate the performance of the developed methodologies against diverse real malware scenarios (static and dynamic analysis, binary and source code files, android and windows environments)

 

GENERAL INFORMATION

This 4 year PhD studentship, potentially funded by the Department for Employment and Learning (DEL), commences on 1 October 2019.

Eligibility for both fees and maintenance depends on the applicants being either an ordinary UK resident or those EU residents who have lived permanently in the UK for the 3 years immediately preceding the start of the studentship. Non UK residents who hold EU residency may also apply but if successful may receive fees only.

Applicants should apply electronically through the Queen’s online application portal at: https://dap.qub.ac.uk/portal/

Further information available at:  http://www.qub.ac.uk/schools/eeecs/StudyattheSchool/PhDProgrammes

 

Contact Details:

Principal Supervisor(s):  Jesus Martinez del Rincon

Email:  j.martinez-del-rincon@qub.ac.uk

Telephone: +44 (0)28 9097 1779