RSS and Spatial Correlation Based Approach to Detect and Localize Spoofing Attackers in Wireless Networks

Abstract: Wireless networks are vulnerable to attacks due to shared medium and mobility besides resource constrained nature. Spoofing attacks are one kind of attacks that can be launched easily on wireless networks. Cryptography can be used to overcome this problem. However, conventional security approaches cannot be directly used with wireless networks as they cause overhead. Without depending on cryptography, recently Yang et al. proposed a method that uses spatial information to detect spoofing attacks, find number of attackers involved in the attacks and also localizing the adversaries. Spatial correlation of received signal strength is used in order to achieve this. SVM and cluster based mechanism are used to know number of attackers and improve accuracy of detection. In this paper we build a prototype network application that simulates the wireless networks. It is used to demonstrate the detection and localization of spoofing attackers. The empirical results reveal that the proposed application is capable of detecting attacks besides localizing attackers.

Keywords: Wireless networks, spoofing attacks, localization, and attack detection