Español(Spanish Formal International)English (United Kingdom)
There are no translations available.

Semantic Web-enabled Network Management.

The growing complexity, heterogeneity and dynamism inherent in emerging telecommunications networks, distributed systems and advanced information and communication services, as well as their increased criticality and strategic importance in the networked economy, calls for the adoption of increasingly more sophisticated technologies for their management, coordination and integration to assure adequate levels of functionality, performance and reliability.

Of the available technologies, those associated with the autonomous agent-based computation paradigm are precisely the ones that are better accepted for conceiving new techniques for developing management solutions with a higher level of automation, greater potential for interoperability within open environments and better capabilities of cooperation. Autonomous agent technology and, particularly, Multi-Agent Systems provide in this respect a series of new and exciting possibilities in the field of network operations and management, such as formal semantic-level knowledge representation, automatic reasoning and learning capabilities, high-level communication languages and protocols, frameworks for automated negotiation, goal-driven proactive behavior or rational decision making.

The formalisms used in management information modeling and representation are closely related to the capabilities of automation, interoperation and cooperation of the management solutions developed on their basis. The success of the process of incorporating autonomous agents, capable of reasoning and dynamically integrating knowledge and services, as an enabling technology for new management solutions, largely depends on the evolution of the information models of existing management architectures towards explicit declarative-type semantic models, equipped with a solid formal basis, that can capture the semantics of the management information models, as well as their formal specification, communication and automatic reasoning about these models. Knowledge Representation and Conceptual Modeling are the fields of Artificial Intelligence that have progressed most in this respect. However, they have had hardly any impact on any of the management information models built to date.

Considering the advances achieved in the field of Knowledge Representation by the international research community, the strategy followed for building the existing management information models should be reconsidered and the possibility of including techniques related to the field of Knowledge Representation should be examined, as should the benefits of such a decision. In this project, we demonstrate the adequacy of the use of Description Logics and the Web Ontology Language OWL for formally defining the structure and constraints of management information in the context of the information model of a management architecture. This model determines the modelling approach and notation used to describe the managed elements, which includes their identification, structure, behavior and relations to other elements.

Common Information Model (CIM) is the chosen information model. CIM is the standard formalism for modeling management information developed by the Distributed Management Task Force (DMTF) in the context of its WBEM proposal, designed to provide a conceptual view of the managed environment. There is widespread agreement on the need to provide CIM diagrams with precise semantics that can be used to establish a common understanding of the formal meaning of the CIM metamodel constructs used for the purpose of enabling interoperation and cooperation. This point has been repeatedly recognized by the DMTF since a keynote address presented at the IEEE Policy 2003 Conference . Although there are proposals for formalizing structural UML diagrams that are easily adaptable to CIM diagrams, none of these proposals amounts to a solid foundation for the development of automatic reasoning techniques based on algorithms that are sound and complete with respect to the semantics.

The first goal of this project is to propose the inclusion of formal knowledge representation techniques, based on DLs, in CIM-based conceptual modeling. The proposal is carried out as a CIM metamodel level mapping to a highly expressive subset of DLs. The aim is to be able to automatically reason about the management information models conceptualized by CIM both in the design phase (to verify formal properties of the models, such as their satisfiability, extract logical implications from and detect inconsistencies or redundancies in the models) and at run time, through the use of rational agents that are able to exploit the DL-OWL expressions of CIM models and their instances as domain ontologies in their deduction, coordination and action processes. To achieve this latter aim, the project contemplates the use of the Semantic Web ontology language OWL for XML-based representation and exchange of the CIM models previously formalized by means of DLs. This latter point amounts to a significant advance over the use of the MOF (Managed Object Format) textual specification language or CIM/XML mapping proposed by the DMTF.

The second goal is to develop a set of CASE tools (known as the CIMOnt Tool Framework ) for visual ontologies modeling which are able to take advantage of the mapping. These tools will enhance the visual development of CIM models, the formalization in DL of these models, and their OWL specification. This framework is being developed in the context of the Nesmarq Project .


SEMANA Project



  • Javier Soriano
  • Genoveva López

Research Assistants

  • Rafael Fernández
© CoNWeT Lab. 2008-2009.