Evaluation of an Intrusion Detection System for Routing Attacks in Wireless Self-organised Networks

Wireless Sensor Networks (WSNs) arebecoming increasingly popular, and very useful in militaryapplications and environmental monitoring. However,security is a major challenge for WSNs because they areusually setup in unprotected environments. Our goal in thisstudy is to simulate an Intrusion Detection System (IDS)that monitors the WSN and report intrusions accurately andeffectively. We have thus simulated an IDS that usesanomaly-based technique to monitor traffic pattern on thenetwork following a fixed-width clustering algorithm. Oursimulation is based on the sensor network simulationpackage by the Naval Research laboratory (NRL). Toevaluate the IDS, we simulated a sensor network,investigated it with the presence of phenomenon, andextended it to generate denial of service attacks. We haveused the phenomenon contribution to generate a realistictraffic pattern for accurate evaluation of protocols, andcompared it to the traditional method of using only cbrtraffic, which is usually been employed by mostresearchers. We further adapted the IDS into this simulatednetwork, and our results show that the selected IDS hasdetection rate of over 90% with a very low false positiverate of less than 1%. We obtained this by configuring everynode to independently monitor detect and report intrusions.