Localized performance-guided reconfiguration for distributed sensor networks

Abstract

Sensor networks that can reconfigure to extend their useful operating life while guaranteeing coverage performance are of interest. This paper describes the use of localized performance-guided reconfiguration (PGR) for reconfiguring a distributed sensor network in a target localization scenario using binary sensor data. If a sensor in a cluster becomes unavailable due to a loss of battery power or other reasons, PGR activates redundant sensors as necessary to allow the network to achieve the desired coverage performance. For a simple demonstration of localized PGR, we compare the performance of a multicluster network using localized PGR, a nearest-neighbor reconfiguration approach, and an approach where no reconfiguration of the field is done. Without PGR, coverage performance degraded as sensors were lost. With PGR, every time a sensor failed, the respective cluster was reconfigured to maintain coverage. In this way, clusters made local reconfiguration decisions to achieve the global field coverage performance goal.