Decentralized data processing and management in Smart Grid via gossiping

In this paper, we propose a model for a peer-to-peer database system architecture where each server is aware of the power system state vector, and that can provide a scalable architecture for the Smart Grid. The key new component is a decentralized gossip-based state estimation (G-SE) algorithm included as a stored database procedure, which constantly updates via network diffusion, the global state information vector in each server of the distributed database system (DDBS). In addition to provide global situation awareness, state aware DDBS can attain communication and storage efficiency first, by using the temporal and spatial correlation of the state to compress it and, second, by storing the state as side information to store measurement residuals. Thanks to the G-SE primitive DDBS are more resilient to network attaks, can reduce network usage and streamline query processing for Smart Grid applications.