Um algoritmo distribuído para eleição de líderes de clusters semânticos em redes de sensores sem fio

Abstract

Semantic Clustering is a recent technique for saving energy in wireless sensor networks. Its mechanism of action consists in dividing the network into groups (clusters) formed by nodes semantically related and at least one leader node, which acts as the bridge between the internal nodes to its cluster and sink node. By accumulating collector role semantic, node leader tends to deplete their energy faster than others. Therefore, it is necessary to use mechanisms for electing leaders semantic to prolong your lifetime. Our hypothesis is that an effective choice of leaders allows an increase in the lifetime of the network. To test it, we start from a previous work of the authors of this article [Rocha et al. 2012] and we propose an algorithm for electing leaders in distributed way based on a fuzzy inference engine. The inputs of the inference engine are the residual energy of nodes and their received signal strength indicator (RSSI). Simulation results confirm our hypothesis, since the algorithm provides (i) an improvement of 17.4% in relation to another proposal of the related literature, and (ii) a gain of 68.8% over the time life of the network's original work.