【24h】

Buffer Sizing in 802.11 Wireless Mesh Networks

机译:802.11无线网状网络中的缓冲区大小

获取原文

摘要

We analyze the problem of buffer sizing for TCP flows in 802.11-based Wireless Mesh Networks. Our objective is to maintain high network utilization while providing low queueing delays. The problem is complicated by the time-varying capacity of the wireless channel as well as the random access mechanism of 802.11 MAC protocol. While arbitrarily large buffers can maintain high network utilization, this results in large queueing delays. Such delays may affect TCP stability characteristics, and also increase queueing delays for other flows (including real-time flows) sharing the buffer. In this paper we propose sizing link buffers collectively for a set of nodes within mutual interference range called the ''collision domain''. We aim to provide a buffer just large enough to saturate the available capacity of the bottleneck collision domain that limits the carrying capacity of the network. This neighborhood buffer is distributed over multiple nodes that constitute the network bottleneck; a transmission by any of these nodes fully utilizes the available spectral resource for the duration of the transmission. We show that sizing routing buffers collectively for this bottleneck allows us to have small buffers (as low as 2 - 3 packets) at individual nodes without any significant loss in network utilization. We propose heuristics to determine these buffer sizes in WMNs. Our results show that we can reduce the end-to-end delays by 6× to 10× at the cost of losing roughly 5% of the network capacity achievable with large buffers.
机译:我们分析了基于802.11的无线网状网络中TCP流的缓冲区大小问题。我们的目标是保持较高的网络利用率,同时提供较低的排队延迟。无线信道的时变容量以及802.11 MAC协议的随机访问机制使问题变得复杂。尽管任意大的缓冲区可以维持较高的网络利用率,但这会导致较大的排队延迟。这样的延迟可能会影响TCP稳定性,并且还会增加共享该缓冲区的其他流(包括实时流)的排队延迟。在本文中,我们建议为相互干扰范围内的一组节点(称为“冲突域”)共同确定链路缓冲区的大小。我们旨在提供一个足够大的缓冲区,以使瓶颈冲突域的可用容量达到饱和,从而限​​制了网络的承载能力。该邻域缓冲区分布在构成网络瓶颈的多个节点上。这些节点中的任何一个节点进行的传输在传输期间将充分利用可用的频谱资源。我们显示出,针对此瓶颈共同确定路由缓冲区的大小可以使我们在各个节点上具有较小的缓冲区(低至2-3个数据包),而不会在网络利用率方面造成任何重大损失。我们提出启发式方法来确定WMN中的这些缓冲区大小。我们的结果表明,我们可以将端到端的延迟减少6倍至10倍,其代价是损失了大缓冲区可实现的大约5%的网络容量。

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