首页> 外文OA文献 >Void-creating algorithms for fixed packet length in OPS/OBS
【2h】

Void-creating algorithms for fixed packet length in OPS/OBS

机译:用于Ops / OBs中固定分组长度的空洞创建算法

代理获取
本网站仅为用户提供外文OA文献查询和代理获取服务,本网站没有原文。下单后我们将采用程序或人工为您竭诚获取高质量的原文,但由于OA文献来源多样且变更频繁,仍可能出现获取不到、文献不完整或与标题不符等情况,如果获取不到我们将提供退款服务。请知悉。
获取外文期刊封面目录资料

摘要

As demand for bandwidth keeps increasing, the optical backbone capacity will have to be used more efficiently in future networks. In order to meet this requirement, both optical packet and optical burst switching are proposed as future network techniques. To resolve packet contention in the used optical switches, Fiber Delay Lines and wavelength converters are used to reschedule the contending packets. This rescheduling process is mastered by means of a scheduling algorithm, typically designed for minimal packet loss in case of a finite buffer. Although, depending on the intentions, a wide variety of scheduling algorithms exists, they can be split up in two big categories: void-filling and non-void-filling algorithms. In contrast to the latter, the former allow packets to be scheduled before already scheduled packets, in this way increasing performance but also the computational complexity. In case the packet lengths are fixed and equal to the granularity, the void-filling algorithms behave as non-void-filling algorithms. This is because only voids smaller than the packet length occur, which cannot be filled. We therefore propose an algorithm that selectively creates larger voids, creating voids only when they will likely be filled. Results obtained by Monte Carlo simulation for a single-wavelength buffer show that these algorithms enable significant improvement in both packet loss probability and packet delay. Moreover, based on an intuitive reasoning, an approximation of the threshold controlling void creation is proposed. All this opens opportunities to implement the void creation in more complex settings such as a multi-wavelength buffer.
机译:随着对带宽的需求不断增长,在未来的网络中必须更加有效地使用光骨干网容量。为了满足该要求,提出了光分组交换和光突发交换两者作为未来的网络技术。为了解决所使用的光交换机中的数据包争用,光纤延迟线和波长转换器用于重新安排竞争数据包。该重新调度过程通过调度算法来掌握,该算法通常设计为在有限缓冲区的情况下将数据包丢失降至最低。尽管根据意图,存在各种各样的调度算法,但是它们可以分为两大类:无效填充算法和非无效填充算法。与后者相反,前者允许在已调度的数据包之前调度数据包,这样既提高了性能,又提高了计算复杂度。在分组长度是固定的并且等于粒度的情况下,空隙填充算法表现为非空隙填充算法。这是因为仅会出现小于数据包长度的空白,而无法填充。因此,我们提出了一种算法,该算法有选择地创建较大的空洞,仅在可能会填充空洞时才创建空洞。通过蒙特卡罗模拟获得的单波长缓冲器的结果表明,这些算法可以显着改善丢包率和丢包延迟。此外,基于直观的推理,提出了控制空隙产生的阈值的近似值。所有这些为在更复杂的设置(例如多波长缓冲区)中实现空隙创建提供了机会。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利

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