A fully-decentralized semantic mechanism for autonomous wireless sensor nodes

Semantic sensor neighborhood has been used to organize nodes into clusters in wireless sensor networks. Semantic clusters are self-adaptable according to information collected/gathered from sensor nodes and collaboratively processed. In this paper, we show that semantic clustering based on fully-decentralized semantic neighborhood mechanisms provides an energy-efficient solution, thus contributing to increase the autonomy of sensors. Our results show that fully-decentralized semantic clustering outperforms partially decentralized semantic clustering algorithms besides traditional clustering algorithms regarding the energy consumed to establish and maintain the clusters.