An Optimal Sensor Architecture for Wi-Fi Intrusion Detection

This paper presents a novel approach to identifying illegitimate nodes in wireless networks. It demonstrates that, given a sufficiently dense sensor network, it is able to discriminate all network nodes based solely on signal strength datasets. Further, it demonstrates that within the dense sensor network, only a small subset of available sensors are required to effectively discriminate between nodes, and that this subset can be readily identified. Finally, it proposes that duplicated network nodes may be identified by unsupervised, real-time analysis of signal strength datasets.