Machine Leaning DNS data analysis for Automated MaliciousDomain Classification

Aaron Ridley, Robert Abbas, Ponnappan Ponnurangam

Research output: Working paper

Abstract

Due to the exponential growth of Internet of Things (IoT) devices in recent years, combined with the often under-securing of such devices, the rise of botnets targeting these devices for recruitment has significantly increased. Once compromised, a part of these botnets can be used to orchestrate devastating distributed denial of service (DDoS) attacks. Unfortunately, imposing standards on such devices is a challenging task because the manufacturing process is cheap and rapid mass. Therefore, another solution to the increasing threat of DDoS attacks must be formed. This paper details a responsive machine learning based solution to detect outgoing malicious traffic such as various DDoS attacks, as well as providing a proof of concept and an interesting sneak peak into the emerging world of software defined security solutions.
Keywords:
Original languageEnglish
Publication statusIn preparation - 20 Feb 2019

Keywords

  • IoT security
  • DDoS mitigation
  • machine learning

Fingerprint Dive into the research topics of 'Machine Leaning DNS data analysis for Automated MaliciousDomain Classification'. Together they form a unique fingerprint.

  • Cite this