Interoperable Semantic and Syntactic Service Discovery for Ambient Computing Environments
Résumé
The inherent heterogeneity of ambient computing environments and their constant evolution requires middleware platforms to manage networked components designed, developed and deployed independently. Such management must also be efficient to cater for resource-constrained devices and highly dynamic situations due to the spontaneous appearance and disappearance of networked resources. For service discovery protocols (SDP), one of the main functions of service-oriented architectures (SOA), the efficiency of the matching of syntactic service descriptions is most often opposed to the fullness of the semantic approach. As part of the PLASTIC middleware for ambient computing environments, we present in this paper an interoperable service matching and ranking platform, which is able to process service descriptions from both semantic and syntactic service description languages. To that end, we define a generic, modular description language able to record service functional properties, potentially extended with semantic annotations. An evaluation of the prototype implementation of our platform demonstrates that multi-protocols service matching supporting various levels of expressiveness can be achieved in ambient computing environments.