Data labeling for integrity in SCADA systems

Supervisory Control and Data Acquisition (SCADA) systems attain important data for critical infrastructure monitoring and control. The data is also used by various high level applications in automation engineering, decision support, operation planning and scheduling, and measurement and accounting. Due to the increasing security threats and system failures, SCADA data may be compromised, corrupted, or produced by malfunctioned devices/applications. Ensuring a high level of data integrity is critical for the operations of SCADA systems and the supported business functions. In this paper, a data labeling approach is proposed to ensure data integrity in SCADA systems. Essentially, a label records the component information of a data item as well as how the data is generated. This information is used to assess how damage to a component data could result in damage to compound data items and, more importantly, how to detect such damage and correct any affected data. For any compromised data that has been recovered, its label must be updated, called relabeling. We define relabeling strategies and communication models to propagate component updates to all the affected data items to ensure that they can be corrected, thus to maintain data integrity. To our best knowledge, our work is the first of its kind to specifically address data integrity issues in SCADA systems.