Lawrence Berkeley National Laboratory
SPARCS: Stream-processing architecture applied in real-time cyber-physical security
- Author(s): Gentz, R
- Peisert, S
- Boverhof, J
- Gunter, D
- et al.
Published Web Locationhttps://doi.org/10.1109/eScience.2019.00028
© 2019 IEEE. In this paper, we showcase a complete, end-To-end, fault tolerant, bandwidth and latency optimized architecture for real time utilization of data from multiple sources that allows the collection, transport, storage, processing, and display of both raw data and analytics. This architecture can be applied for a wide variety of applications ranging from automation/control to monitoring and security. We propose a practical, hierarchical design that allows easy addition and reconfiguration of software and hardware components, while utilizing local processing of data at sensor or field site ('fog computing') level to reduce latency and upstream bandwidth requirements. The system supports multiple fail-safe mechanisms to guarantee the delivery of sensor data. We describe the application of this architecture to cyber-physical security (CPS) by supporting security monitoring of an electric distribution grid, through the collection and analysis of distribution-grid level phasor measurement unit (PMU) data, as well as Supervisory Control And Data Acquisition (SCADA) communication in the control area network.