DETECTION AND LOCALIZATION OF MULTIPLE SPOOFING ATTACKERS FOR MOBILE WIRELESS NETWORKS

Abstract

The openness nature of wireless networks allows adversaries to easily launch variety of spoofing attacks and causes havoc in network performance. Recent approaches used Received Signal Strength (RSS) traces, which only detect spoofing attacks in mobile wireless networks. However, it is not always desirable to use these methods as RSS values fluctuate significantly over time due to distance, noise and interference. In this paper, we discusses a novel approach, Mobile spOofing attack DEtection and Localization in WIireless Networks (MODELWIN) system, which exploits location information about nodes to detect identity-based spoofing attacks in mobile wireless networks. Also, this approach determines the number of attackers who used the same node identity to masquerade as legitimate device. Moreover, multiple adversaries can be localized accurately. By eliminating attackers the proposed system enhances network performance. We have evaluated our technique through simulation using an 802.11 (WiFi) network and an 802.15.4 (Zigbee) networks. The results prove that MODELWIN can detect spoofing attacks with a very high detection rate and localize adversaries accurately.