Autoencoder Based Anomaly Detection for SCADA Networks

Supervisory control and data acquisition (SCADA) systems are industrial that used to monitor critical infrastructures such as airports, transport, health, public services of national importance. These cyber physical systems, which increasingly integrated with networks internet things devices. However, this results in a larger attack surface for threats, making it important identify thwart cyber-attacks by detecting anomalous network traffic patterns. Compared other techniques, well known patterns, machine learning can also detect new evolving threats. Autoencoders type neural generates compressed representation its input through reconstruction loss inputs help data. This paper proposes the use autoencoders unsupervised anomaly-based intrusion detection using an appropriate differentiating threshold from distribution demonstrate improvements compared techniques SCADA gas pipeline dataset.