Autonomous Decentralized Flow Control in High-Speed Networks with Inhomogeneous Configurations
Autonomous Decentralized Flow Control in High-Speed Networks with Inhomogeneous Configurations
复制标题
非均匀配置高速网络中的自主分散流量控制
DOI：
--
复制
发表时间：
2004
影响因子：
0.7
通讯作者：
S. Kuribayashi
中科院分区：
文献类型：
--
作者：
C. Takano;Masaki Aida;S. Kuribayashi
最近计算机通信的增长导致对高速骨干网络的需求增加。在这样的高速网络中，时间敏感的流量控制机制所采用的原则应该是自治分散控制的原则。在这种机制中，网络中的每个节点仅根据直接可用的本地信息来管理其本地流量，尽管希望每个节点做出的单独决策能够带来整个网络的高性能。在我们以前的研究中，我们已经调查了本地数据包流的行为和全局性能时，一个节点是拥塞，并提出了扩散型流量控制模型。然而，由于我们在评估中使用了一个简单而均匀的网络模型，因此结果不能推广。在本文中，我们提出了一个扩展的扩散型流量控制模型，以便将其应用到网络的非均匀配置。我们展示了两种情况下的仿真结果：不同的传播延迟和多个瓶颈。这两个结果都表明，所提出的扩散型流动条件
SUMMARY Recent growth in computer communications has led to an increased requirement for high-speed backbone networks. In such highspeed networks, the principle adopted for a time-sensitive flow control mechanism should be that of autonomous decentralized control. In this mechanism, each node in a network manages its local traffic flow only on the basis of the local information directly available to it, although it is desirable that the individual decisions made at each node lead to high performance of the network as a whole. In our previous studies, we have investigated the behavior of local packet flows and the global performance achieved when a node is congested, and proposed the diffusion-type flow control model. However, since we used a simple and homogeneous network model in the evaluation, the results cannot be generalized. In this paper, we propose an extension of the diffusion-type flow control model in order to apply it to networks with inhomogeneous configurations. We show simulation results for two cases: different propagation delays and multiple bottlenecks. Both results show that the proposed diffusion-type flow con