A demand - assignment algorithm based on a Markov modulated 3 chain prediction model for satellite bandwidth allocation 4 Francesco Delli

7 Abstract This article deals with the problem of the 8 design of a control-based demand-assignment algorithm for 9 a satellite access network using a Markov modulated chain 10 traffic prediction model. The objective is to guarantee a 11 target Quality of Service (QoS) to Internet traffic, while 12 efficiently exploiting the air interface. The proposed algo13 rithm is in charge of dynamically partitioning the uplink 14 bandwidth capacity in a satellite spotbeam among the in15 progress connections. Such partition is performed aiming at 16 matching the QoS requirements of each connection and 17 maximizing the satellite bandwidth exploitation. A closed18 loop Control Theory approach is adopted to efficiently 19 tackle the problem of the delay between bandwidth 20 requests and bandwidth assignments, while minimizing the 21 signaling overhead caused by control messages. The 22 algorithm efficiently copes with both the satellite propa23 gation delay and the delays inherent in the periodic nature 24 of the bandwidth request mechanism. The proposed 25 demand-assignment algorithm and Markov chain traffic 26 prediction model are shown to improve the overall satellite 27 network performance through extensive simulation 28 experiments. 29 30