A Novel Free-Rider Detection Scheme for IEEE 802.11 Wireless Networks

In this paper, we propose a novel free-rider detection scheme for IEEE 802.11 wireless networks. In our proposed scheme, various characteristics extracted from different layers in the TCP/IP stack are exploited to detect a free-rider. Specifically, these characteristics are combined and applied to a simple but powerful probabilistic-based classifier. To the best of our knowledge, our scheme is the first to target the detection of a ``mercurial´´ free-rider. In addition, it does not require additional hardware support or changes to the current IEEE 802.11 standards so as to implement our proposed scheme. A comprehensive evaluation study is conducted to examine the performance of our proposed scheme. The evaluation results demonstrate that the proposed scheme is able to effectively detect mercurial free-riders under different network settings.