Achieving Optimized Traffic Sharing over Equal-Cost-Multi-Paths Using LRU-based Caching with Counting Scheme

LIN Wei; LIU Bin; TANG Yi

您当前的位置：

首页 >

文章列表页 >

Achieving Optimized Traffic Sharing over Equal-Cost-Multi-Paths Using LRU-based Caching with Counting Scheme

更新时间：2025-07-16

- Achieving Optimized Traffic Sharing over Equal-Cost-Multi-Paths Using LRU-based Caching with Counting Scheme
- Acta Electronica Sinica Vol. 36, Issue 1, Pages: 32-38(2008)
- 作者机构：
  
  1. 清华大学深圳研究生院,广东,深圳,518055
  2. 香港城市大学,香港
  3. 清华大学计算机科学与技术系,北京,100084
  4. 清华大学深圳研究生院广东深圳,518055
  5. 香港城市大学香港
  6. 清华大学计算机科学与技术系北京,100084
- 作者简介：
- 基金信息：
- DOI：
  CLC： TP393.05
- Published：2008
- 稿件说明：
移动端阅览
LIN Wei, LIU Bin, TANG Yi. Achieving Optimized Traffic Sharing over Equal-Cost-Multi-Paths Using LRU-based Caching with Counting Scheme[J]. Acta Electronica Sinica, 2008, 36(1): 32-38.
DOI：

LIN Wei, LIU Bin, TANG Yi. Achieving Optimized Traffic Sharing over Equal-Cost-Multi-Paths Using LRU-based Caching with Counting Scheme[J]. Acta Electronica Sinica, 2008, 36(1): 32-38. DOI：

摘要

为了减少网络拥塞并充分利用链路带宽

当在转发节点与目的子网间存在有多条等价路径(ECMPs)时

流量负载应该在ECMPs间均衡分配

并且属于同一个TCP流的IP分组应该按照相同顺序到达目的主机.本文提出了一种基于LRU(Least Recently Used Algorithm) Cache和计数统计的算法.该算法通过为每条ECMP分配一个计数器

利用计数统计从而考虑到了IP分组的长度差异.使用相对计数以及对某些情况增加约束条件解决了计数器溢出问题.UDP分组只需要作为调节负载均衡的流量.更进一步

对于去往同一目的子网的不同主机的TCP流的时延差异被转化为cache中的表项失效的时间长度差.仿真实验表明

当ECMPs间的时延差不显著的情况下

只需要很小的存储空间

且每次cache查找只需要一个时钟周期

负载均衡接近最优

此时只有2%的分组出现乱序.

Abstract

In order to reduce network congestion and fully use link bandwidth

when there are Equal-cost-multi-paths (ECMPs) between a forwarding node and a destination subnet

traffic load should be balanced among ECMPs and packets of the same TCP flow should reach destination host in the same order.An algorithm called LRU-based caching with counting (LCC) is proposed.Packet length differentiation is considered to achieve load balance by adapting a counter for each ECMP

and counter overflow is solved by relative counting and restrictions.UDP packets only need to be concerned to achieve load balance.Furthermore

flow delay differentiation forwarding to different hosts of the same destination subnet is transformed to entries in cache invalided time period difference.Simulation shows that when delay differentiation among ECMPs is not significant

storage requirement is small

only one cycle is needed for each cache lookup

load balance is near optimal

and only 2% of packets are out of order.

关键词

Keywords

references

Views

1198

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

The Research on Power Controlling Policies for Instruction Cache with Architecture Level Methods

The Research on Leakage Power Controlling Policies for On-Chip L2 Cache

A High Performance Embedded SoC:TengYue-1

Pre-Visiting Tag and Keeping Way History to Reduce Power in Instruction Cache

A Design of the Disk Array Controller

Related Author

ZHANG Min-xuan

ZHOU Hong-wei

ZHANG Min-xuan

QI Shu-bo

OU Guo-dong

ZHOU Hong-wei

LU Hong-yi

WANG Zhi-ying

Related Institution

Department of Computer Science,National University of Defense Technology.

Tianjin Navigation Instrument Research Institute.

Department of Computer ScienceNational University of Defense Technology.ChangshaHunan 410073China

Tianjin Navigation Instrument Research Institute.Tianjin 300131China

School of Computer Science,National University of Defense Technology

⁰