Hunting Trojan Horses

  • Authors:
  • Micha Moffie;Winnie Cheng;David Kaeli;Qin Zhao

  • Affiliations:
  • Northeastern University, Boston, MA;Massachusetts Institute of Technology, Cambridge, MA;Northeastern University, Boston, MA;National University of Singapore, Singapore

  • Venue:
  • Proceedings of the 1st workshop on Architectural and system support for improving software dependability
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

HTH (Hunting Trojan Horses) is a security framework developed for detecting difficult types of intrusions. HTH is intended as a complement to anti-virus software in that it targets unknown and zero-day Trojan Horses and Backdoors. In order to accurately identify these types of attacks HTH utilizes runtime information available during execution. The information collected includes fine-grained information flow, program execution flow and resources used.In this paper we present Harrier, an Application Security Monitor at the heart of our HTH framework. Harrier is an efficient run-time monitor that dynamically collects execution-related data. Harrier is capable of collecting information across different abstraction levels including architectural, system and library APIs. To date, Harrier is 3-4 times faster than comparable information flow tracking systems.Using the collected information, Harrier allows for accurate identification of abnormal program behavior. Preliminary results show a good detection rate with a low rate of false positives.