A Data Stream-Based, Integrative Approach to Reliable and Easily Manageable Real Time Environmental Monitoring

Abstract

Real time environmental monitoring (hereafter, RTEM) is crucial for observing and studying dynamic, rare, or abrupt phenomena in the environment. Conventional RTEM systems are developed in a domain-specific way and not based on a well-defined distributed system model. Therefore, it is challenging to develop such RTEM systems, to extend them to support new requirements, and to integrate them with other application systems. In this paper, we raise challenging issues in RTEM, propose a generic distributed system model for RTEM to address those issues, and present the Galilee middleware system for RTEM based on the model. The distributed system model is based on the concept of data streams. It is intended to simplify and to conceptualize the design of RTEM systems. Therefore, the simple data stream-based distributed system model facilitates the development of general RTEM middleware system. The Galilee middleware system for RTEM is based on the distributed system model and designed to support various RTEM applications in a reliable and easily manageable way. In this paper, we also demonstrate and evaluate the current prototype implementation of the Galilee system with those datasets obtained from the previous sensor-based water quality monitoring project for Lake Soyang from 2011 to 2012.