RatBot: anti-enumeration peer-to-peer botnets

  • Authors:
  • Guanhua Yan;Songqing Chen;Stephan Eidenbenz

  • Affiliations:
  • Information Sciences, Los Alamos National Laboratory;Department of Computer Science, George Mason University;Information Sciences, Los Alamos National Laboratory

  • Venue:
  • ISC'11 Proceedings of the 14th international conference on Information security
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

As evidenced by the recent botnet turf war between SpyEye and Zeus, the cyber space has been witnessing an increasing number of battles or wars involving botnets among different groups, organizations, or even countries. One important aspect of a cyber war is accurately estimating the attack capacity of the enemy. Particularly, each party in a botnet war would be interested in knowing how many compromised machines his adversaries possess. Towards this end, a technique often adopted is to infiltrate into an adversary's botnet and enumerate observed bots through active crawling or passive monitoring methods. In this work, we study potential tactics that a botnet can deploy to protect itself from being enumerated. More specifically, we are interested in how a botnet owner can bluff the botnet size in order to intimidate the adversary, gain media attention, or win a contract. We introduce RatBot, a P2P botnet that is able to defeat existing botnet enumeration methods. The key idea of RatBot is the existence of a fraction of bots that are indistinguishable from their fake identities. RatBot prevents adversaries from inferring its size even after its executables are fully exposed. To study the practical feasibility of RatBot, we implement it based on KAD, and use large-scale high-fidelity simulation to quantify the estimation errors under diverse settings. The results show that a naive enumeration technique can significantly overestimate the sizes of P2P botnets. We further present a few countermeasures that can potentially defeat RatBot's anti-enumeration scheme.