Networking and Data Management
From Grid Optics
The future power grid is expected to deploy 100 million smart sensors and meters with various data transfer speeds. The resulting data volume can be at peta scale (1015) over a one-year time period. Power grid planners and policy makers need to have tools to quickly retrieve and analyze these volumes of data to make real-time decisions.
This focus area is addressing the gaps in networking and real-time data management by developing advanced algorithms, techniques, and software tools. These tools and techniques will deliver solid outcomes that support the future power grid's ability to:
- identify, filter, and reduce data based on conditions and locations to ensure real-time performance,
- define scientific principles, requirements and resources to enable large-scale information network modeling and simulation environments; enabling the evaluation of performance and scalability challenges under real-time constraints,
- develop operational sensor prototypes that allow adaptive autonomous operation supporting new distributed control paradigms, and
- address and develop capabilities for secure real-time data transfer (data in motion) and create new methods that ensure integrity, privacy and confidentiality of devices through new cryptographic and non-cryptographic management techniques.
To tackle these questions, are expanding the size and scope of the EIOC to include a flexible and reconfigurable network infrastructure design laboratory, which will validate the models, frameworks, algorithms and tools being developed within the FPGI.
Furthermore, we have developed a data management middleware that advances the state of the art and will be optimized for use in the power grid. The data management middleware will be flexible and scalable to allow data sources and consumers to be added easily and dynamically in response to changes in the power grid. This research will provide a solid foundation to deploy next generation real-time data management systems for the power grid.
We are focusing on simplifying the deployment challenges of large-scale power grid networks by addressing configuration and routing aspects relevant to efficiency and reliability. PNNL is performing research on extending multi-topology routing protocols to support richly connected mesh network topologies and enable multi-path, multi-criteria routing in wired and wireless networks. Another field of research in this focus area is new routing protocols in order to prevent traffic diversion (false route injection) and denial of service attacks. Lastly, by tackling operational complexity associated with public key infrastructures, certificate revocation, identity and key management, we will improve cyber security.
List of Projects:
- GridOPTICS Power Networking, Equipment, and Technology (powerNET) Testbed
- VOLTTRON An Intelligent Agent Platform for the Smart Grid
- Next Generation Network Simulations for Power System Applications
- GridOPTICS: A Software Framework for Power System Operations and Planning
- Scalable Sensor Data Management Middleware
- Multi-Resolution Data Model and Directed Data Reduction, Reconstruction, and Aggregation
- Framework for Network Co-Simulation Video (FNCS)
- GridOPTICS Software System Video (GOSS)