首页> 外文会议>Communication networks and distributed systems modeling and simulation conference(CNDS'99) >Performance Modeling of a Multistage Buffered ATM Switch with Bursty Traffic
【24h】

Performance Modeling of a Multistage Buffered ATM Switch with Bursty Traffic

机译:具有突发流量的多级缓冲ATM交换机的性能建模

获取原文
获取原文并翻译 | 示例

摘要

In this paper, we develop a new approximate decomposition method for the performance analysis of a multistage buffered ATM switch with bursty traffic. The restrictions of uniform taffic model and destination in dependence assumptions are not necessary. After aggregating the traffic at each crossbar by utilizing the results in splitting and merging processes of bursty traffic, each of the crossbars is isolated from its switches so that it can be analyzed separately. In this way an equivalent buffered flow line is recursively formed and it is much easier to be analyzed. A group of nonlinear equations are formed based on the conservation-of-flow and solved for each buffer parameters. The performance metrics of interest can be evalauted, such as average throughput, buffer occupancy, packet normalized bandwidth and delay. Comparing to the existing methods, the proposed method can be applied to the following cases: 1). even or bursty packet traffic, like Poisson arrivals, MEG distributions, and Bernoulli arrivals. 2) fixed or general packet legnth distribution. 3) arbitrary buffer capacity, small or relatively large size. 4) large switching networks, because the analytical complexity doesn't increase as the the network size. The effectiveness of the proposed method has been verified by many numerical examples.
机译:在本文中,我们开发了一种新的近似分解方法,用于突发流量的多级缓冲ATM交换机的性能分析。不必依赖依赖假设来限制统一流量模型和目的地。通过利用突发流量的拆分和合并过程中的结果汇总每个交叉开关处的通信量之后,每个交叉开关都与其交换机隔离,以便可以对其进行单独分析。以此方式,等效缓冲流线被递归形成,并且更容易分析。基于流量守恒形成一组非线性方程,并针对每个缓冲区参数求解。可以评估感兴趣的性能指标,例如平均吞吐量,缓冲区占用率,数据包归一化带宽和延迟。与现有方法相比,所提出的方法可以应用于以下情况:1)。偶数或突发数据包流量,例如泊松到达,MEG分布和伯努利到达。 2)固定或一般的分组长度分配。 3)任意缓冲区容量,较小或相对较大的大小。 4)大型交换网络,因为分析复杂度不会随着网络规模的增加而增加。许多数值例子验证了该方法的有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号