An efficient real-time misbehavior detector using non-parametric cumulative sum (CUSUM) test inWLAN

摘要

The CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) based wireless protocol ensures that each node in the network gets a reasonably fair share (FS) of the network channel. In the distributed environment of wireless network, a malicious node may deliberately select a smaller backoff timer to gain an unfair share of the network channel, thus disrupting the network throughput. This nature is called misbehavior in WLAN. There are two challenges in misbehavior detection: (i) unknown misbehavior detection, where a malicious node initially behaves as a normal node and at any time later may change its backoff timer to a random small value. (ii) Realtime misbehavior detection, where the misbehavior needs to be detected in real time. The existing solutions either cannot address either of the issues or require modifications to the 802.11 protocols. The objective of the project is to quickly detect the misbehavior in real time without any prior knowledge of the statistics of the misbehavior and then isolate the malicious node as soon as possible to prevent it from bringing more harm to the network. We design a real-time backoff misbehavior detector, termed as Fair Share detector (FS detector) that makes use of a nonparametric cumulative sum (CUSUM) test to quickly find a malicious node. The system guarantees performance in terms of average false positive rate, average detection delay, and missed detection ratio under a detection delay constraint. The project is widely used in current programmable and reconfigurable wireless network devices which are easily available.