EASY: Efficient semAntic Service discoverY in pervasive computing environments with QoS and context support

Pervasive computing environments are populated with networked software and hardware resources providing various functionalities that are abstracted, thanks to the Service Oriented Architecture paradigm, as services. Within these environments, service discovery enabled by service discovery protocols (SDPs) is a critical functionality for establishing ad hoc associations between service providers and service requesters. Furthermore, the dynamics, the openness and the usercentric vision aimed at by the pervasive computing paradigm call for solutions that enable rich, semantic, contextand QoS-aware service discovery. Although the Semantic Web paradigm envisions to achieve such support, current solutions are hardly deployable in the pervasive environment due to the costly underlying semantic reasoning with ontologies. In this article, we present EASY to support efficient, semantic, contextand QoS-aware service discovery on top of existing SDPs. EASY provides EASY-L, a language for semantic specification of functional and non-functional service properties, as well as EASY-M, a corresponding set of conformance relations. Furthermore, EASY provides solutions to efficiently assess conformance between service capabilities. These solutions are based on an efficient encoding technique, as well as on an efficient organization of service repositories (caches), which enables both fast service advertising and discovery. Experimental results show that the deployment of EASY on top of an existing SDP, namely Ariadne, enhancing it only with slight changes to EASY-Ariadne, enables rich semantic, contextand QoS-aware service discovery, which furthermore performs better than the classical, rigid, syntactic matching, and improves the scalability of Ariadne.