A Semantic Fragrance Trail based Scalable Service Discovery in MANET

Abstract

The scale of MANET is expanding with the proliferation of mobile devices, and the number of services in MANET also increases. It raises the importance of scalability and semantic searching capability. However, existing service discovery protocols targeting MANET have been focused on either only scalability or only semantic matching. Some recent researches are concerning the scalable semantic service discovery, but they take into account only one semantic, which is not sufficient to find relevant services in case there are many different service instances in the same service type. In this paper, we propose a new scalable semantic service discovery utilizing rich semantic information based on a scheme, so called semantic fragrance trail (SFT). For advertising services, Service BBF (SBBF) is propagated in a scalable way. For discovery, query messages are routed following the most relevant semantic fragrance (SF) by comparing SSFs from neighbor nodes to the queries using our proposed equation comparing SF similarity. We prove that the proposed equation can compare SSF and semantic fragrance of a query (QSF) to find the most relevant service through evaluation.