A linear system analysis of RED

The random early detection (RED) algorithm proposed by Floyd and Jacobson [IEEE/ACM Trans. Networking 1 (4) (1993) 397] is a simple and quite effective mechanism for active queue management and as a result has been widely implemented in Internet routers. While RED certainly performs better than drop-tail, its parameters are difficult to tune to give good performance under different congestion scenarios. Also, studies have shown that RED does not stabilize the queue lengths in routers, because its equilibrium queue length strongly depends on the number of active TCP connections. As a result numerous variants of RED have been proposed to work around the performance problems of RED. In this paper, using inferences drawn from a linear systems analysis, we provide an insight into why RED is difficult to tune. The same linear systems analysis is used to show that the DRED algorithm proposed in [Computer 36 (2–3) (2001) 203; Comput. Commun. 24 (12) 2001 1170] overcomes the performance limitations of RED [IEEE/ACM Trans. Network. 1 (4) (1993) 397].     

Managing Internet routers congested links with a Kohonen-RED queue

The behaviour of the TCP AIMD algorithm is known to cause queue length oscillations when congestion occurs at a router output link. Indeed, due to these queueing variations, end-to-end applications experience large delay jitter. Many studies have proposed efficient active queue management (AQM) mechanisms in order to reduce queue oscillations and stabilize the queue length. These AQM attempt to improve the random early detection (RED) model. Unfortunately, these enhancements do not react in a similar manner for various network conditions and are strongly sensitive to their initial setting parameters. Although this paper proposes a solution to overcome the difficulties of configuring the RED parameters by using a Kohonen neural network model; another goal of this study is to investigate whether cognitive intelligence could be placed in the core network to solve such stability problem. In our context, we use results from the neural network area to demonstrate that our proposal, named Kohonen-RED (KRED), enables a stable queue length without complex parameters setting or passive measurements to obtain a correct configuration.     

TCP/RED离散模型分析及参数设置

随机早期检测(RED)队列管理算法的参数设置主要依赖于使用者的经验,并且其相关理论指导存在一定缺陷。为此,提出一种全面设置RED参数的方法。采用数学期望方法,分析TCP/RED拥塞控制机制离散模型的稳定状态行为,推导网络稳定状态期望值与RED参数的解析关系式。仿真结果表明,该方法可以提供准确有效的RED参数设置。     

TCP New Veno:一种改进的TCP拥塞控制机制

TCP Veno协议通过对慢启动、拥塞避免和快速恢复的修改,改进了传统的TCP Reno的性能。然而,TCP的不公平性的问题仍然有待于解决。参与竞争的TCP流之间的不平衡可能造成某些通信源垄断队列空间。例如,当长RTT和短RTT流共存时,网络流量会逐渐集中于短RTT链路上。提出了一种新的TCP拥塞控制机制——TCP New Veno。其基本思路是通过引入带宽预测和动态窗口变化的思想进一步改进TCP Veno的性能,并导出其数学模型。数学分析和仿真实验都证明,改进后的算法在保证吞吐量的基础上提高了原算法的公平性。     

统计平衡状态下的RED建模和仿真

RED作为取代 TD的 Internet路由器队列管理策略 ,虽然已经得到了广泛的认同 ,但到目前为止其参数设置主要还是依赖于使用者的经验 ,这严重影响了它的大范围实现和部署 .通过均值分析方法对 RED在统计平衡时的动态特性进行了建模 ,并得出了在多个 TCP流竞争一条瓶颈链路的情况下 ,平均队长和 RED参数之间的解析关系 .仿真结果表明 ,该模型可以提供准确和有效的 RED参数设置方法 .     

一种针对TCP动态队列管理的控制算法

描述了一种新的TCP/IP动态队列管理控制算法,对于传统的随机早期检测法(RED)来说,它的主要目标之一就是稳定路由器队列的长度,然而它实现此目标并不是很成功,主要因为它在平衡队列长度的过程中很强地依赖了动态TCP链接数。而新的动态控制算法则使用了一种简单的控制方法,当路由器缓冲区即将出现拥塞时,它能够根据当前路由器缓冲区负载概率来随机地实施包丢弃。该算法能够很好地稳定路由器缓冲区的队列占用数,同时,在实现过程中并不用评估动态TCP链接数以及分析网络流的状况。所给出的一个实验模型表明,该控制算法是有效可行的。     

RED算法的数学模型研究

RED算法能够及时预测网络拥塞的到来,并同过标记避免网络拥塞,同时还解决了TCP全局同步的问题.RED算法对参数过于敏感一直是研究的主要问题.建立RED算法的数学模型,从数学角度分析RED算法的原理以及工作过程.对数学模型的静态分析、参数分析研究各个参数在算法中所起的作用以及其影响因子;给出如何设定参数才能使RED算法达到更好的性能.     

MANET环境下的TCP Veno性能研究

采用NS2对运行在MANET环境下的TCP Veno的性能进行仿真测试,仿真结果表明,在有背景流、随机丢包且存在拥塞的 MANET网络中,TCP Veno的性能优于TCP Reno,且背景流越大、达到拥塞的时间越短、随机丢包越大,TCP Veno的优越性就越明显。但2种协议的兼容性较差,导致数据包的重传率增大。     

Estimation of the variances of TCP/RED using stochastic differential equation

By using the stochastic differential equation (SDE), a model is built to describe the system behavior, especially the queue length behavior, of TCP/RED (transmission control protocol with random early detection) on a single bottleneck network. Then, not only the static equilibrium point of the system is calculated, but also its variance is estimated. After calculating the variance of the queue length in the router, the authors show how such random oscillation affects the maximum connection number that the link admits while keeping low dropping probability, and how this relation is influenced by other network parameters. This model can be applied to the case where only homogeneous TCP flows exist, as well as where TCP and UDP (user data protocol) flows coexist. The simulations by NS‐2 show that our deductive method is effective. Copyright © 2011 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society     

Bounds estimation and practical stability of AIMD/RED systems with time delays

The Additive Increase and Multiplicative Decrease (AIMD) congestion control algorithm of TCP deployed in the end systems and the Random Early Detection (RED) queue management scheme deployed in the intermediate systems contribute to Internet stability and integrity. Previous research based on the fluid-flow model analysis indicated that, with feedback delays, the TCP/RED system may not be asymptotically stable when the time delays or the bottleneck link capacity becomes large [3]. However, as long as the system operates near its desired equilibrium, small oscillations around the equilibrium are acceptable, and the network performance (in terms of efficiency, loss rate, and delay) is still satisfactory. In this paper, we study the practical stability of AIMD/RED system with feedback delays and with both homogeneous and heterogeneous flows. We obtain theoretical bounds of the flow window size and the RED queue length, as functions of the number of flows, link capacity, RED queue parameters, and AIMD parameters. Numerical results with Matlab and simulation results with NS-2 are given to validate the correctness and demonstrate the tightness of the derived bounds. The analytical and simulation results provide important insights on which system parameters contribute to higher system oscillations and how to set parameters (such as buffer size and queue management parameters) to ensure system efficiency with bounded delay and loss. Our results can also help to predict and control the system performance for Internet with higher data rate links multiplexed with heterogeneous flows with different parameters.     

Dynamic queue level control of TCP/RED systems in AQM routers

One main TCP congestion control objective is, by dynamically adjusting the source window size according to the router queue level, to stabilize the buffer queue length at a given target, thereby achieving predictable queueing delay, reducing packet loss and maximizing link utilization. One difficulty therein is the TCP acknowledging actions will experience a time delay from the router to the source in a TCP system. In this paper, a time-delay control theory is applied to analyze the mechanism of packet-dropping at router and the window-updating in TCP source in TCP congestion control for a TCP/RED dynamic model. We then derive explicit conditions under which the TCP/RED system is asymptotically stable in terms of the instantaneous queue. We discuss the convergence of the buffer queue lengths in the routers. Our results suggest that, if the network parameters satisfy certain conditions, the TCP/RED system is stable and its queue length can converge to any target. We illustrate the theoretical results using ns2 simulations and demonstrate that the network can achieve good performance and converge to the arbitrary target queues.     

TCP Veno——无线环境下的TCP拥塞控制改进机制

在无线网络环境下,TCPReno和TCPVegas受到无线信道干扰和噪声的影响,其性能大幅降低。为克服无线环境下由于随机丢包而产生的性能下降,Fu等提出了一种新的无线网络环境下TCP拥塞控制机制——TCPVeno。TCPVeno综合了TCPReno和TCPVegas的特点,通过对慢启动、拥塞避免和快速恢复算法的改进,使TCP协议在无线网络环境下性能有了较大的改善,并兼顾了协议的公平性、灵活性。     

基于NS2的主动队列管理仿真研究

主动式队列管理技术是IETF为了解决Internet 拥塞控制问题而提出的一种路由器缓存管理技术。该文介绍了目前应用较为广泛的网络仿真器NS-2，对几种主要AQM算法Drop Tail、RED和ARED的性能在基于NS2仿真实验的基础上进行了比较研究，研究的性能包括队列长度、TCP 全局同步问题、连接数对系统稳定性和鲁棒性的影响等；仿真结果表明ARED性能优于Drop Tail和RED算法。     

一种分阶段自适应RED／ECN参数模型

TCP拥塞控制技术是IP网络性能的研究重点之一,目前已经出现了多种改进方案,然而采用这些方案后,TCP连接仍然面临大量的包丢失,尤其是在网络拥塞时。为此IETF提出了IP网络的拥塞控制问题并建议在网络中采用主动队列管理机制。RED／ECN算法是一种主动队列管理算法,它具有很多优点,但算法对其静态参数的依赖性很大,没有充分考虑链路中复用的活动连接数,因而不能很好地适应网络负荷的变化。为此,提出了一种分阶段动态调整RED／ECN算法参数的模型（GARED）,通过仿真实验说明该模型改进了RED／ECN算法的性能,能够有效地降低丢包率,保证网络链路的高利用率。     

基于PID控制器的TCP主动队列管理*

运用反馈控制理论分析了TCP模型的稳定性能,并设计PID控制器以提高该系统的稳定性,采用OPNET对带PID控制器的TCP拥塞控制系统进行了仿真。仿真结果表明,PID控制器比传统的RED具有更快的响应速度、更小的队列长度波动和更大的链路利用率。     

A framework to determine bounds of maximum loss rate parameter of RED queue for next generation routers

Random early detection (RED) is expected to eliminate global synchronization by random active packet drop. Its packet drop probability is decided by the maximum packet drop probability in its drop function, buffer thresholds, and average queue length. It has been observed that for a large number of connections, a small value of the maximum packet drop probability may not eliminate global synchronization. Furthermore, since RED uses four parameters to regulate its performance, it is necessary to relate its maximum drop probability with those parameters. The objective of this paper is to develop a framework for the bounds of the maximum drop probability of RED, based on TCP channel model and traffic characteristics. The value of the maximum drop probability obtained by our model will make RED queue achieve its targeted goals.     

基于测量的TCP拥塞控制的公平性研究

通过分析传统TCP算法的局限性，讨论TCP Vegas、TCPW两种基于源端实时带宽测量拥塞控制算法的原理以及带宽分配的公平性，结合主动队列管理技术，提出一种基于加权缓存区容量分配RED算法．理论分析和仿真实验表明该算法提高了带宽分配的公平性．保持了网络的高吞吐量，并实现服务QoS保证.     

TCP Veno在3G环境下的性能测试

将TCP Veno运行在3G UMTS(WCDMA)环境下,采用NS-2对其性能进行仿真测试.仿真结果表明:在存在背景流、有随机丢包并且存在拥塞的3G网络中,TCP Veno的性能优于TCP Reno,而且背景流越大、达到拥塞的时间越短、随机丢包越大,TCP Veno的优越性会更加明显.     

Proportional loss rate differentiation in a FIFO queue

The relative differentiated service model is one of several models proposed for service differentiation in networks [IEEE Network Sept/Oct (1999) 26]. In this model, an assurance is given that ‘higher classes will be better, or at least no worse than lower classes.’ This paper describes a relative loss rate differentiation scheme based on RED. The scheme is used for differentially dropping packets in a FIFO queue during times of congestion. The main idea is, if packet losses are unavoidable in the FIFO queuing system, then they should be distributed among the different service classes in the queue in inverse proportion to the service price or weight assigned to each class. The simulation studies using TCP traffic show that the scheme is very effective in ensuring relative loss rate differentiation between service classes.