2Systems and Security Department
4Computer Science Department
Institute for Infocomm Research (I2R), Singapore
Columbia University, USA
3Dept. of Comp. Science and Engineering
Cambridge University, UK
University of California San Diego, USA
The growing popularity of wireless networks and mobile devices is starting to attract unwanted attention especially as potential targets for malicious activities reach critical mass. In this study, we try to quantify the threat from large-scale distributed attacks on wireless networks, and, more specifically, wifi networks in densely populated metropolitan areas. We focus on three likely attack scenarios: "wildfire" worms that can spread contagiously over and across wireless LANs, coordinated citywide phishing campaigns based on wireless spoofing, and rogue systems for compromising location privacy in a coordinated fashion. The first attack illustrates how dense wifi deployment may provide opportunities for attackers who want to quickly compromise large numbers of machines. The last two attacks illustrate how botnets can amplify wifi vulnerabilities, and how botnet power is amplified by wireless connectivity.
To quantify these threats, we rely on real-world data extracted from wifi maps of large metropolitan areas in the States and Singapore. Our results suggest that a carefully crafted wireless worm can infect up to 80% of all wifi connected hosts in some metropolitan areas within 20 minutes, and that an attacker can launch phishing attacks or build a tracking system to monitor the location of 10-50% of wireless users in these metropolitan areas with just 1,000 zombies under his control.