Markov decision-based optimisation on bundle size for satellite disruption/delay-tolerant network links
Markov decision-based optimisation on bundle size for satellite disruption/delay-tolerant network links
复制标题
基于马尔可夫决策的卫星中断/延迟容忍网络链路捆绑大小优化
DOI：
10.1049/iet-com.2017.1089
复制
发表时间：
2018
影响因子：
1.6
通讯作者：
Yang ZH
中科院分区：
文献类型：
--
作者：
Li Yue;Wang Ye;Yuan Peng;Yang Zhihua;Yang ZH
在卫星中断/延迟容忍网络中，由于成对轨道节点之间的距离和连接性不断变化，捆绑交付显然受到时变参数（即误码率和传播延迟）的影响。作者提出了一种基于马尔可夫决策的捆绑大小优化方法，可以有效地提高预期的交付时间在一个动态的两跳星间链路。特别是，一组最佳的束大小自适应地选择根据当前的距离相关的信道参数，从而充分利用中间节点的内存。仿真结果验证了所提出的方法在不同条件下的比较。
In a satellite disruption/delay-tolerant network, bundle delivery is obviously affected by time-varying parameters, i.e. bit error rate and propagation latency, due to constantly changing distance and connectivity between paired orbital nodes. The authors proposed a Markov decision-based optimisation approach for bundle size, which could efficiently improve the expected time of delivery over a dynamic two-hop inter-satellite link. In particular, a group of optimal bundle sizes are adaptively selected according to current distance-dependent channel parameters, leading to a full utilisation on intermediate node's memory. The simulation results verified the proposed method under different conditions with comparison.