Research On Cross-layer Scheduling And Resource Allocation Methods In Satellite Internet

With the rapid development of Internet and the explosive growth of trafficapplications, it is an inevitable trend that satellite network becomes a supplement ofthe Internet for its advantage of wide coverage. However traditional satellitenetwork stack design is layered, this layered optimized only at various protocollayers independently, without considering the optimization of the overall networkperformance, the results of the network’s overall performance is not optimal. Thecross-layer framework is of fundamental importance in order to adapt the differentstrategies available at the various layers. Moreover, it not only leads to improvedmultimedia performance over existing satellite networks, but also provides valuableinsights into the design of next-generation algorithms and protocols. Thisdissertation focuses on the satellite network downlink packet cross-layer schedulingalgorithms and cross-layer-aware uplink resource allocation algorithm. In aparticular type of service QoS requirements and system resource constraints forjoint optimization of the network protocol stack to achieve optimal o verallperformance, efficient use of the network resources.We first propose a cross-layer queuing and scheduling architecture architecturefor a QoS-aware transmission scheme. We focus on a NUM to formulate therate/delay balancing when different application flows are feeding the MACscheduler. The utility of each flow is prioritized according to the correspondingvideo quality and delay requirements. Using channel state information in physicallayer, queue information of IP layer, the scheduler make packet joint scheduling inMAC layer, and calculate the various services and appropriate transmission rateback to the application layer of satellites. Through simulation analysis, thecross-layer packet scheduling algorithm achieve delay guarantee for traffic which islatency-critical, while maximizing the satisfaction of other traffic, achievingefficient use of network resources.Furthermore the dissertation makes study on the resource allocation algorithmon satellite return link, we propose to partially decentralize the bandwidth allocationof the DAMA scheme on the return link of a multibeam broadband satellite airinterface. The algorithm uses different network knowledge information(capacityunaware, capacity aware and requests aware) in the pricing, we model each scenarioas a non-cooperative game. The satellite terminals to reach the optimal networkoperating point (Pareto Optimal solution), and thus performing efficientbandwidth-aware requests.