Evaluation of clustering algorithms for DAP placement in wireless smart meter network

Wireless smart meter network has attracted extensive research interests for its capability on intelligent control and monitor the power consumption in residential as well as commercial area. Wireless smart meter network constitutes with a number of smart meters to sense the electricity or gas consumption, data aggregation point (DAP) to collect data from smart meters, and utility center to gather all data from DAPs. In this work, we investigate the clustering and optimal DAP placement problem in the multi-hop smart meter network. Three representative clustering algorithms, namely k-means, self-organizing map, and fuzzy c-means, are evaluated and compared in terms of multi-hop shortest path distance (MSPD), size of clusters and computation complexity. Simulation results indicate that the hard assignment based methods, k-means and self-organizing, achieve similar performance whereas the soft assignment based fuzzy c-means falls behind with longer maximum MSPD and higher complexity.