Congestion pricing for dynamic bandwidth allocation in satellite networks: A game-theoretic approach

The purpose of this work is to provide an enhanced adaptive approach to classic queue-based Bandwidth-on-Demand (BoD) procedures in broadband satellite networks. In queue-based schemes, the controller's objective is to drive the buffer queue length to an appropriate reference queue length, and the efficiency of the control strictly depends on the choice of its reference value. In latest approaches, solutions to adaptively modify the target reference queue length based on a set of network information have been proposed; nevertheless, in such approaches, the satellite terminals should base their control strategies on information which are not available to them, according to the communication standards of state-of-the-art technologies, such as DVB-RCS. The novelty of the presented algorithm is that it overcomes this limit by dynamically changing the reference queue length, based on information actually broadcast to the network terminals; the algorithm is at same time simple and effective, and is based on a game-theoretic approach which benefits of recent advances in congestion pricing theory. The effectiveness of the proposed approach is supported by Opnet® simulations as well as by considerations about the real implementability of the solution.