Date of Award


Publication Type

Master Thesis

Degree Name



Computer Science

First Advisor

Christie Ezeife


Applied sciences




This thesis proposes an Intrusion Detection System, NeuDetect, which applies Neural Network technique to wireless network packets captured through hardware sensors for purposes of real time detection of anomalous packets. To address the problem of high false alarm rate confronted by the current wireless intrusion detection systems, this thesis presents a method of applying the artificial neural networks technique to the wireless network intrusion detection system.

The proposed system solution approach is to find normal and anomalous patterns on preprocessed wireless packet records by comparing them with training data using Back-propagation algorithm. An anomaly score is assigned to each packet by calculating the difference between the output error and threshold. If the anomaly score is positive then the wireless packet is flagged as anomalous and is negative then the packet is flagged as normal. If the anomaly score is zero or close to zero it will be flagged as an unknown attack and will be sent back to training process for re-evaluation.