基于链路负载自适应的主动队列管理算法研究

发布时间：2018-07-17 15:38

【摘要】：随着计算机网络技术的发展，互联网（Internet）规模呈现爆发式的增长、互联网用户也快速增长，IP业务发展迅速，需求呈现多样化，网络流量的急剧增加造成网络拥塞问题日益严重，计算机网络服务质量（QoS）的要求越来越高。为了提高计算机网络服务质量，，基于网络中间节点的拥塞控制机制成为研究的热点，主动队列管理算法（AQM）是基于中间节点拥塞控制机制实现的重要手段，陆续出现了很多关于AQM的算法。本文深入分析了几种经典AQM算法，并总结了各自算法的优缺点。在ARED算法的基础上，为了达到加强队列长度稳定性以及提高链路利用率的目标，提出了一种基于链路负载自适应的改进算法TTL-ARED。ARED算法通过检测平均队列长度来判断网络拥塞的程度，调节最大丢包概率来达到稳定队列长度的效果，但存在网络滞后性，并不能够很好地适应网络中的突发流量。TTL-ARED算法能够根据链路负载的变化情况来判断网络拥塞的程度，达到稳定队列长度的目的，保证了链路高利用率。同时对丢包概率计算函数进行改进优化，能够更好地适应网络拥塞控制，降低分组丢包概率，提升链路的吞吐量。利用网络仿真软件NS2对TTL-ARED算法性能进行了实验仿真，仿真结果表明TTL-ARED算法能够更好地适应不同的网络环境，有效地进行网络拥塞控制，提升了队列长度的稳定性，保证了链路的高吞吐量，具有更好的稳定性和鲁棒性。
[Abstract]:With the development of computer network technology, the scale of Internet (Internet) presents explosive growth, Internet users also grow rapidly, the demand is diversified, the network congestion problem becomes more and more serious due to the rapid increase of network traffic. The quality of service (QoS) of computer network is becoming more and more important. In order to improve the quality of service (QoS) of computer networks, congestion control mechanism based on network intermediate nodes has become a hot topic. Active queue management algorithm (AQM) is an important means to realize congestion control mechanism based on intermediate nodes. There have been a lot of AQM algorithms. In this paper, several classical AQM algorithms are analyzed, and their advantages and disadvantages are summarized. Based on the ARED algorithm, in order to enhance the queue length stability and improve the link utilization, An improved algorithm, TTL-ARED.ARED, based on link load adaptation, is proposed to determine the degree of network congestion by detecting the average queue length and to adjust the maximum packet loss probability to stabilize the queue length. TTL-ARED algorithm can not well adapt to the network burst traffic. TTL-ARED algorithm can judge the degree of network congestion according to the change of link load, achieve the purpose of stabilizing queue length, and ensure high link utilization. At the same time, the packet loss probability calculation function is improved and optimized, which can better adapt to the network congestion control, reduce packet loss probability, and improve the throughput of the link. The performance of TTL-ARED algorithm is simulated by network simulation software NS2. The simulation results show that TTL-ARED algorithm can better adapt to different network environments, effectively control network congestion, and improve the stability of queue length. It ensures the high throughput of the link and has better stability and robustness.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.06

【相似文献】