Semantic-enhanced resource discovery for CoAP-based sensor networks

The integration of knowledge representation and reasoning techniques (originally devised for the Semantic Web) in most common Wireless Sensor Networks (WSNs) protocols can allow to reach higher levels of autonomicity w.r.t. classic network architectures that basically provide only simplistic discovery capabilities. This paper presents a complete Semantic Sensor Network (SSN) framework, supporting a resource discovery based on non-standard inferences. A backward-compatible extension of Constrained Application Protocol (CoAP) has been proposed to support semantic matchmaking for retrieving and ranking resources annotated w.r.t. a reference ontology. Data mining procedures were also exploited to detect high-level events from gathered raw data. A case study on environmental monitoring has been proposed to test the effectiveness of our approach.