A Semi-Supervised Approach for Detection of SCADA Attacks in Gas Pipeline Control Systems

The imperative role played by Supervisory Control And Data Acquisition (SCADA) systems in providing a centralized control for modern infrastructure have made them into one of the most desired targets for malicious attackers owing to its rapid evolution as well as widespread adoption of these systems. To counter these attacks, it is necessary that more robust approaches be adopted. The advent of Machine Learning has shown great potential for its usage along with existing Intrusion Detection Systems (IDS). This paper presents a novel approach to detect malicious behaviour in SCADA data used to control gas pipeline system. As most of the data available in this industry are unsupervised, this paper uses an approach that makes use of a Semi-Supervised Deep Learning architecture- Autoencoder, that is believed to be most suited for this type of tasks. The effectiveness of this deep learning network is due to the fact that it reconstructs the input as the output and in the training process learns only the most important features of normal observations that are representative of the input data; thus malicious data is easily detected due to a high reconstruction error. The proposed algorithm is validated on gas pipeline control system dataset and found to give excellent results in detection.