Abstract—Wireless Sensor Networks are proven highly successful in many areas, including military and security monitoring. In this paper, we propose a method to use the edge–betweenness community detection algorithm to determine clusters and to facilitate in-network data aggregation for these applications. To minimize the cost of determining the clusters, the approach is based on exploiting the topology information from the ad hoc routing protocol. Three different schemes are proposed. (1) A distributed clustering scheme using the OLSR routing protocol. (2) A centralized scheme using OLSR. (3) A centralized scheme using an extension to the DYMO-low routing protocol. All schemes support sensor heterogeneity allowing that different data content can use different routing paths. The paper presents simulation results and an analysis of the cluster generation for each of the schemes. The results show that our method is a simple and effective method to improve scalability and lifetime of r...
Joakim Flathagen, Ovidiu Valentin Drugan, Paal E.