Refine
Year of publication
- 2017 (1) (remove)
Computing Aggregates on Autonomous, Self-organizing Multi-Agent System: Application "Smart Grid"
(2017)
Decentralized data aggregation plays an important role in
estimating the state of the smart grid, allowing the determination of
meaningful system-wide measures (such as the current power generation,
consumption, etc.) to balance the power in the grid environment. Data
aggregation is often practicable if the aggregation is performed effectively. However, many existing approaches are lacking in terms of fault-tolerance. We present an approach to construct a robust self-organizing
overlay by exploiting the heterogeneous characteristics of the nodes and
interlinking the most reliable nodes to form an stable unstructured overlay. The network structure can recover from random state perturbations
in finite time and tolerates substantial message loss. Our approach is
inspired from biological and sociological self-organizing mechanisms.