Ilka Reis,  Gilberto Câmara,  Renato Assunção,  Antônio Miguel Vieira Monteiro. 

Distributed clustering algorithms play an important role in energyefficient data collection proposals for geosensor networks. The available dataaware clustering algorithms build clusters around nodes’ representatives, which represent their associated nodes individually. We propose to build clusters around clusters’ representatives, which are able to produce data summaries that are better estimates to their associated nodes’ data. We present the Distributed Data-aware Representative Clustering (DARC) algorithm. We have concluded the DARC builds more homogeneous clusters and produce data summaries that estimate the nodes’ data with the smallest error, if compared with the usual data-aware clustering proposals.

http://www.lbd.dcc.ufmg.br/colecoes/wtr/2008/009.pdf

Processando consulta.