Attack Classification Analysis of IoT Network via Deep Learning Approach

Abstract

A variety of attacks in the transportation layer of IoT network seeks for a detection and prevention
mechanism such as intrusion detection systems (IDSs). Anomaly detection is one of the most demanding
task in IDSs. It requires a robust classifier model which is able to detect different kinds of
attacks intelligently. This paper addresses deep neural network for classifying attacks in IoT network.
The performance of the proposed method is evaluated on the three novel benchmarking datasets in
wired and wireless network environment, i.e. UNSW-NB15, CIDDS-001, and GPRS. Furthermore,
deep neural network combined with grid search strategy are utilized to obtain the best parameter settings
for each dataset. The experimental results demonstrate the effectiveness of our approach using
deep neural network in terms of accuracy, precision, recall and false alarm rate.