Applying and Validating AI-Based QoS Profiling for NGN User Terminals |
This paper addresses the identification of NGN (Next Generation Networks) user terminals experiencing similar QoS (Quality of Service) conditions. This approach aims on the reduction of network traffic resulting from comprehensive QoS monitoring. An ART 2 ANN (Adaptive Resonance Theory 2 Artifical Neural Network) has been evaluated for the comparison and classification of sequences of consecutive jitter (delay variation) values experienced by packets of simultaneous multimedia over IP data streams. Further on, a procedure is introduced supporting the classification process through automated result validation.
Weber F, Fuhrmann W, Trick U, Bleimann U, Ghita BV