An intelligent agent system for managing heterogeneous sensors in dispersed and disparate wireless sensor network

Abstract

Wireless sensor networks (WSNs) employ middleware solutions to coordinate their sensors and provide web services for managing data. Adding/configuring a new or existing sensor requires modification to the middleware and web service. To integrate hundreds of sensors with different capabilities, a system needs to support all the encodings, models and services for registering, tasking and querying sensors. We present an intelligent agent, which provides automatic semantic-based registration/configuration in a large-scale sensor observation system. The agent can react to any changes internally/externally made (i.e., adding a new sensor or sending a task request to sensors/support systems). The agent makes a 'smart' decision on which system function to use by integrating a semantic representation of sensor network data including middleware and web service specifications and applying logical rules to the knowledge-base. The agent's operation is demonstrated using two real-world systems represented in RDF using a domain ontology that extends the W3C SSN-XG ontology.