An event-based component model for sensor networks: a case study for river monitoring
Jó Ueyama, Danny Hughes, Nelson Matthys, Wouter Horré, Sam Michiels, Christophe Huygens, Wouter Joosen · Lirias · 2010
This paper examines an event-based component model for wireless sensor networks called LooCI (Loosely-coupled Component Infrastructure) and discusses an application of LooCI for river monitoring in creeks that flow through São Carlos. Our sensor network application which is based on LooCI monitors a flood, pollution and human tampering and warns potential stakeholders whenever they are at risk (e.g. of floods). Our pollution monitoring system is used to inspect rivers and issues a warning whenever the condition of water reaches an unacceptable level. In addition, our depth sensor monitors rivers and is used to predict potential floodings. Finally, our three dimensional accelerometer detects any improper vibration caused by tampering with the deployed node. Our component model for sensor networking environments has a novel loosely-coupled binding model which is particularly designed to support intermittent connections between the components. This reflects the nature of wireless sensor networks that are often subject to unreliable networking.