A Semantically Distributed Approach to Map IP Traffic Measurements to a Standardized Ontology

摘要

Traffic monitoring in IP networks is a key issue for operators to guarantee Service Level Agreement both to their clients and with regards to other connectivity providers. Thus, having efficient solutions for traffic measurement and monitoring supports a good deal of their business and it is essential to fair development of Internet. However, even if service management is well recognized, QoS strategies must evolve from circuit switching technological framework towards next generation networks and convergent services concepts. Standardizing IP traffic measurement is a requirement for interoperable service aware management systems upon which future Internet business would be based. A few projects have recently tackled the task of building rich infrastructures to provide IP traffic measurements. The European project MOMENT approach combines SOA and semantic search concepts: a mediator between clients and measurement tools has been designed in order to offer integrated access to the infrastructures, regardless their specific details, with the possibility of achieving complex queries. Pervasiveness of ontologies has been used for various purposes in the project. As such, one ontology deals traffic measurement data, another one describes metadata that is used instead of data for practical reasons, a third one focuses on anonymization required for ethical (and legal) restrictions and the last one describes general concepts from the field. This paper outlines the role of these ontologies and presents the process to achieve them from a set of traffic measurement databases as well as the integration of specific modules in the mediator to achieve the semantic queries.