Anmerkungen:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Beschreibung:
Modern Industrial Control System (ICS) represent a wide variety of networked infrastructure connected to the physical world. Depending on the application, these control systems are termed as Process Control Systems (PCS), Supervisory Control and Data Acquisition (SCADA) systems, Distributed Control Systems (DCS) or Cyber Physical Systems (CPS). Nowadays, the internet has been evolved as a universal communication platform in many domains, including ICS. The major technical background of the latest industrial revolution (Industrie 4.0 or Smart Factories) is the introduction of internet technologies into the industry making the field devices, machines, plants and factories connected to a network. As ICS is designed for reliability; but security especially against cyber threats is also a critical need. Despite several measures, every day a new attack against the ICS is being identified. Therefore, a proper measure is necessary to identify those novel attacks and ensure security. Cybersecurity through detection of malicious activities in ICS by efficiently configuring the deep learning algorithms is the main research foci of this thesis. Through research, the cyber-attacks on ICS can be broadly classified as network attacks or injection attacks. In order to develop the deep learning-based cybersecurity, a proper dataset providing the possible attacks on an ICS is necessary. For network attacks, different datasets do exist. Out of them, NSL-KDD is popularly used by many researchers and is selected for the development of Intrusion Detection System (IDS) in ICS for network attacks. As no proper dataset exists for injection attacks, a dataset for injection attacks is simulated using the data from process control plant in the institute. In order to identify the novel or unknown attack, anomaly-based intrusion detection technique is developed using different deep learning algorithms for classification of normal to anomalous behaviour and a proof-of concept was implemented. The implementations are done in MATLAB using ...