Wireless intrusion detection system using fingerprinting

Abstract

Wireless network is the network which is easy to deploy and very easy to access that network and that network is user friendly. The main reason behind of getting popular is because it provide benefits, like as easy to installation, flexibility, mobility, scalability and reduced cost-of-ownership. But drawback in these wireless networks is that it doesn't provide security as much as required, due to that user faces attacks of various types which are damageable to user information. One of the serious attack is Identity based attacks which steals the identity of some other user in that network and performed some other attack. The available present security tools to detect such these identity(spoofed MAC) based attacks are quite limited. In this proposed work a new technique is developed for detecting masquerade(identity) attacks or spoofed MAC attack exploited in 802.11 wireless network. Current methods of device fingerprinting includes only probe request packets fingerprinting, which results in large amount of false positive. In our proposed work fingerprint is created on basis of three frames which are required in three section of connectivity phase and that frames are probe request frame, authentication frame and association frame. Time differences between consecutive frames are take into consideration and on the basis of that fingerprint is created of different device. In this proposed technique cross-correlation method is used to estimate the signals similarity in terms of time lagging to each other. Those signals are captured by different devices. Stored signature of actual device and captured signal of transmitting device is compared using this technique and after that result analysis, identification of device is done.