CLUSTERING AND NEURAL MODELING FOR PERFORMANCE EVALUATION OF MOBILE COMMUNICATION NETWORKS

Dimitar Radev - Izabella Lokshina

   In this paper, we present a core network model of universal mobile telecommunication system with calls that belong to one of four service classes and arrive randomly. Arriving calls are granted service based on specific service class, required maximum and minimum bandwidth, and available network resources. Performance of priority-based dynamic capacity allocation, suitable for the wireless ATM system is analyzed. Scheduling of the ATM cell transmission in each time division multiple access frame for the uplink is based on a priority scheme. Blocking probability and throughput parameters for bandwidth sharing policy are considered, and partial overlap link is implemented. The clustering procedure for the performance analysis of the mobile communication networks and the blocking probability and throughput measurements are introduced as Markov reward models enhanced with vector quantification and neural modeling. The optimal link occupancy probability distribution is determined using neural network that was trained on the base of Kohonen rules. Simulation and numerical results are shown.

Keywords: mobile communication networks, Markov reward models, clustering, and neural modelling

[full-paper]