College of Sustainable System Sciences;
Osaka Prefecture University;
Ltd.;
Osaka University;
Fuji Electric Co.;
Graduate School of Information Science and Technology;
School of Knowledge and Information Systems;
Improving the scalability and robustness of wireless sensor networks is an important task, and much research on self-organization has been conducted toward this end. However, desired behavior is not yet guaranteed in much larger networks based on pure self-organization. In this article, we propose a controlled potential-based routing protocol implementing a novel controlled self-organization scheme that also allows for external control. The scheme obtains close-to-optimal network behavior by this external control which controls a part of nodes in the network. We show that global traffic flow can be controlled through simulation experiments with a multi-sink sensor network. For example, traffic loads can be equalized among heteroge-neously distributed sink nodes, and load balancing among the relay nodes based on remaining energy can bring an approximate four times extension of network lifetime. The proposed method is furthermore robust to message loss and resilient to failure of the sink node.