【作者】 陈丽娟； 吴玺； 黄翌；

【Author】 CHEN Li-juan1 WU-Xi2 HUANG Yi3 1 Department of Electronic Engineering and Information Science,Hefei University of Technology, Hefei, 230009 2 School Of Computer & Information, ,Hefei University of Technology, Hefei, 230009 3 CETC 38, Hefei, 23009

【机构】 合肥工业大学仪器科学与光电工程学院； 合肥工业大学计算机与信息学院； 中电集团38所；

【摘要】 分析研究了用于主动队列管理的随机早期检测(RED)算法和BLUE算法,指出RED算法对于业务负载变化敏感,而BLUE算法平均队列长度变化波动大,针对它们的优缺点,提出使用平滑滤波算法管理到达分组。通过仿真实验结果比较,平滑滤波算法能够更好地减小链路拥塞时到达网关分组的丢失率,提高网络吞吐量,使资源分配效率有较大提高,并且对于不同的业务负载,性能非常稳定。更多还原

【Abstract】 After analyzing the Random Early Detection (RED) algorithm and the BLUE algorithm for active queue management, this paper points out that the RED algorithm is sensitive to the traffic loads, but by using the Blue algorithm the length of average queue is fluctuant much more. For those reason, smooth filter algorithm is provided here. After doing some simulations, comparing the results, by using the smooth filter algorithm, packets loss rate can drop down obviously, network throughput is improved more, resources distribution becomes more efficiently. Also, to the different type of traffic load, the performance of smooth filter algorithm is very stable.更多还原