Correlation-based Clustering of Telecommunication Equipment in Smart Grid

Telecommunication equipment in smart grids is expanding and producing a considerable amount of daily data that must be monitored efficiently. The availability of multiple sensors in the network generates an abundance of data, making network resource management a challenging task. In this paper, correlation-based clustering of smart grid equipment with similar traffic patterns is proposed as a solution to simplify resource management and monitoring tasks. The availability time-series data was used to study the correlation between equipment in the smart grid, and then the correlated equipment was divided into clusters. These clusters created groupings that were unknown by the management center yet aligned with the telecommunication company architecture. The proposed method simplifies the resource management process and reduces personnel time and cost, by informing network technicians if one piece of equipment is unavailable, then the whole cluster is expected to be unavailable.