An algorithm to coordinate missions in wireless sensor networks

Abstract

This paper presents an algorithm to allocate suitable nodes of a Wireless Sensor Network (WSN) to accomplish a given task (called mission). This algorithm is part of the Grubmi middleware. A mission is inserted in an arbitrary point of the network, being disseminated to all nodes by the middleware. Then, a suitable set of nodes should be selected to execute the jobs described by the mission. The network is heterogeneous, therefore, the nodes' characteristics must match the mission's requirements. Since a centralized algorithm with global information about the network is prohibitive due to the costs, we proposed in this paper a probabilistic approach to solve this problem. Two versions of the heuristic are described: the basic and the extended one. For a scenario with mixed placement of different nodes, both variants, in average, presented results which were 87% of the optimal one. For scenarios with regions of nodes with similar characteristics, the basic heuristic had lower performance whereas the extended one could maintain the same behavior.