论文部分内容阅读
确定性延迟容忍网络中,节点间的接触往往表现出一定的重复性和规律性.现有的确定性路由算法通常根据网络全局或部分的先验知识构建网络演化图,将路由选择过程转换为利用经典路由算法求解最短路径的问题.然而,这类算法中消息的转发一般采用预先计算好的路径,缺乏对网络状态的自适应性.网络流量可能集中于部分的活跃节点,造成这部分节点过度的资源消耗,从而导致网络拥塞.提出一种基于节点介数的拥塞感知路由算法,该算法通过网络拓扑的时空演化图计算出节点间延时开销最小的多条备选路径,同时引入节点介数来指示节点的负载情况.在转发路径的选择过程中,结合路径的延迟开销和介数值以不同的概率从备选路径集合中选择实际转发路径.仿真结果表明该算法有效地减少了网络负载不均造成的局部拥塞现象,提高了网络消息的交付性能.
Deterministic delay tolerant networks often show some repeatability and regularity of contacts between nodes.Present deterministic routing algorithms usually construct the network evolution graph according to the global or partial prior knowledge of the network and convert the routing process into Use the classical routing algorithm to solve the shortest path problem.However, the forwarding of messages in this kind of algorithm generally adopts pre-calculated path, lack of adaptability to network status.Network traffic may be concentrated in part of the active node, causing this part of the node Excessive resource consumption, which leads to network congestion.This paper proposes a congestion-aware routing algorithm based on node number, which uses the space-time evolution graph of network topology to calculate multiple candidate paths with the least delay overhead between nodes, In the process of selecting the forwarding path, the actual forwarding path is selected from the set of candidate paths with different probabilities according to the delay overhead and the meso-value of the path.The simulation results show that the algorithm effectively reduces the network Local congestion caused by uneven load improves the delivery performance of network messages.