Towards an Intelligent System for Self-Correction of QoS in IMS Networks

The IMS Network QoS Architecture approach for monitoring and supervision services in the IMS offers innovative solutions for QoS management in real time. These solutions enable the integration of automatic correcting mechanisms for QoS deterioration, depending on network conditions and customer importance. However, the correction operation is costly in terms of resources and execution time. Face to this problem, the migration to an intelligent system based on reasoning based cases will enable the exploitation of recorded solutions for similar degradation cases. In this context, we propose a methodology for structuring both cases and recorded solutions in such a way as to minimize the research cost. This methodology can be held in two stages, the clustering for grouping similar cases and the classification for backup structure generation of event.