Detecção de intrusos por observação em redes de sensores sem fio

Abstract

This work proposes an Intrusion Detection System for Wireless sensor networks (WSNs) using the information available in the base station. WSNs are ad hoc networks comprised mainly of small sensor nodes with limited resources, and can be used for monitoring areas of interest. Applications range from battlefield reconnaissance to environmental protection. In some of these applications the network will certainly be attacked, and an intrusion detection system needs to be in place to detect the presence of intruders. Given their distinguishing characteristics, WSNs require tailored intrusion detection systems. However, their design and implementation are challenging because of the resource constraints. The proposed system does not require changed in the analyzed network. We propose an information model for intrusion detection in the RSSF and an extensible arquitecture, which can be adapted for different comportment models and different detection strategies. The experiments show that the detection rate was above 80% and the number of false positives were 405 in 4000 analyzed events.