No abstract

No abstract

International audienceData aggregation is a promising method for conservingenergy in Wireless Sensor Networks (WSNs). The mostefficient approach to aggregation is organizing the networkinto clusters. However, this approach which achieves goodperformance in periodic data collecting applications is provedto be inappropriate for event-driven applications, such as realtimedata gathering. In fact, in this case maintaining clustersturned to be very expensive in term of message overhead. Todeal with this problem, we propose a clustering scheme suitablefor event-driven applications. We show that if the structure isadapted to the occurrence of events, structured approaches canprovide good results in real-time WSNs. Simulation results showthat our approach performs well both in aggregation gain andenergy consumption

No abstract