WiFi network usage is increased rapidly these days while the number of attacks in WiFi network are growing as well. Intrusion Detection System (IDS) is one of the popular defense mechanisms that often uses e.g., machine learning algorithms in order to detect both known and unknown attacks in a particular network. We leverage an unsupervised deep learning approach, so called Stacked Auto Encoder (SAE) as feature extraction scheme. Feature extraction by SAE can reduce the complexity of original features of the dataset. While regression layer with softmax activation function is implemented as supervised classification. In this paper, we test our proposed IDS using AWID dataset which is one of comprehensive WiFi network traces from real network. Our experiments show that our proposed IDS can outperform the previous work by Kolias et al. In addition, we provide several suggestions in order to made our proposed IDS reach an optimum result.
[1]
Pascal Vincent,et al.
Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion
,
2010,
J. Mach. Learn. Res..
[2]
Georgios Kambourakis,et al.
Intrusion Detection in 802.11 Networks: Empirical Evaluation of Threats and a Public Dataset
,
2016,
IEEE Communications Surveys & Tutorials.
[3]
Jeffrey M. Voas,et al.
Learning Internet-of-Things Security "Hands-On"
,
2016,
IEEE Security & Privacy.
[4]
Benjamin Schrauwen,et al.
Sign Language Recognition Using Convolutional Neural Networks
,
2014,
ECCV Workshops.
[5]
Thomas Hofmann,et al.
Greedy Layer-Wise Training of Deep Networks
,
2007
.
[6]
Kwangjo Kim,et al.
Detecting Impersonation Attack in WiFi Networks Using Deep Learning Approach
,
2016,
WISA.