Interactive, Visual-Aided Tools to Analyze Malware Behavior

Abstract

Malicious software attacks can disrupt information systems, violating security principles of availability, confidentiality and integrity. Attackers use malware to gain control, steal data, keep access and cover traces left on the compromised systems. The dynamic analysis of malware is useful to obtain an execution trace that can be used to assess the extent of an attack, to do incident response and to point to adequate counter-measures. An analysis of the captured malware can provide analysts with information about its behavior, allowing them to review the malicious actions performed during its execution on the target. The behavioral data gathered during the analysis consists of filesystem and network activity traces; a security analyst would have a hard time sieving through a maze of textual event data in search of relevant information. We present a behavioral event visualization framework that allows for an easier realization of the malicious chain of events and for quickly spotting interesting actions performed during a security compromise. Also, we analyzed more than 400 malware samples from different families and showed that they can be classified based on their visual signature. Finally, we distribute one of our tools to be freely used by the community.