A Comprehensive Ontology for Internet of Things (COIoT)

In the Internet of Things (IoT), interoperability among heterogeneous entities is an important issue. Most existing solutions offer to address the interoperability by proposing a horizontal service platform to enable communication interoperability between machines. Without realizing the true meaning of data, interoperability is restricted to addressing protocols and message formats. Semantics enable interoperability in the true sense by allowing applications to be autonomic. However, most of the current ontologies for IoT mainly focus on modelling resources, services, and location information and/or define completely new ontologies defeating the purpose of using them for interoperability. Introduction of new ontologies is causing more fragmentation in building a consistent and comprehensive knowledge base for IoT. Careful reuse of existing specialized ontologies can help in developing an interoperable semantic IoT system. Comprehensive Ontology for IoT (COIoT) reuses core concepts from existing ontologies like SSN, iot-lifecycle, PowerOnt, OWL-S, GeoNames, DUL and complements concepts like policy, context, services and monitoring. Using COIoT, it is demonstrated that an interoperable semantic knowledge base for IoT environments can be built and used.

I. Introduction

Internet of Things is rooted in its diversity of the devices, protocols; because of this the information originating from these devices is also naturally different. Heterogeneity coupled with the openness in the domain and further absence of standards have collectively resulted in syntactic and semantic interoperability issues. The syntactic interoperability is getting much of the attention and much of the research has happened in this area. AllJoyn [1], IoTvity [2], AndroidThings [3], SensorML[4] etc. have invested all their energy in developing message formats, protocols, and infrastructure for driving the Things around us. The uptake of these standards are hence driven by the consortia and is restrictive in true sense of interoperability. Also, most of these solutions are focused on achieving interoperability in communication leaving out the crucial issues related to interpretation of data. Applications trying to harness this data require more than just the message format and protocols to be interoperable, they must attribute proper meaning to the data they exchange; that is where semantic interoperability plays a key role. Considering that IoT is increasingly gaining traction with a projection of about 20 billion devices to be connected by 2020 [5] it is crucial that applications and application developers be provided with a concrete way of deciphering the meaning of data. Building large scale collaborative IoT frameworks have been increasingly difficult due to interoperability issues between the IoT deployments. Semantic technologies enable interoperability in the application layer by introducing common vocabularies and interoperable representation of the data. Many ontologies ranging from specific domains like e-health, m-health, transportation, tourism, agriculture etc. to very generic ones like W3C SSN, oneM2M, IoT-O, IoT-lite. Different applications are pushing standards targeted at application domains such as smart-meters standards developed by IEC or IEEE (EN 13757, IEEE 1888–2011, etc.).