Traffic management for TCP/IP over satellite ATM networks

Several Ka-band satellite systems have been proposed that will use ATM technology to seamlessly transport Internet traffic. The ATM UBR, GFR, and ABR service categories have been designed for data. However, several studies have reported poor TCP performance over satellite ATM networks. We describe techniques to improve TCP performance over satellite ATM networks. We first discuss the various design options available for TCP end systems, IP-ATM edge devices, as well as ATM switches for long-latency connections. We discuss buffer management policies, guaranteed rate services, and the virtual source/virtual destination option in ATM. We present a comparison of ATM service categories for TCP transport over satellite links. The main goal of this article is to discuss design and performance issues for the transport of TCP over UBR, GFR, and ABR services for satellite ATM networks     

A new ATM adaptation layer for TCP/IP over wireless ATM networks

This paper describes the design and performance of a new ATM adaptation layer protocol (AAL‐T) for improving TCP performance over wireless ATM networks. The wireless links are characterized by higher error rates and burstier error patterns in comparison with the fiber links for which ATM was introduced in the beginning. Since the low performance of TCP over wireless ATM networks is mainly due to the fact that TCP always responds to all packet losses by congestion control, the key idea in the design is to push the error control portion of TCP to the AAL layer so that TCP is only responsible for congestion control. The AAL‐T is based on a novel and reliable ARQ mechanism to support quality‐critical TCP traffic over wireless ATM networks. The proposed AAL protocol has been validated using the OPNET tool with the simulated wireless ATM network. The simulation results show that the AAL‐T provides higher throughput for TCP over wireless ATM networks compared to the existing approach of TCP with AAL 5. This revised version was published online in July 2006 with corrections to the Cover Date.     

TCP/IPoverATM的性能研究

 本文讨论将传统的TCP拥塞控制机制和CIPOA用于采用ER(FB)算法和EFCI(CCR)算法的ATM网络时可用比特率业务(ABR)和未确定比特率业务(UBR)的性能.仿真结果表明,无论在缓冲区的需求、带宽分配的公平性、吞吐量和链路利用率方面ABR业务的TCP性能均明显优于UBR业务.对于较简单的网络模型ER(FB)算法的TCP性能优于EFCI(CCR)算法,更优于传统的EFCI算法.     

An empirical study of TCP/IP performance over FR/ATM service interworking

This article outlines some basic performance characteristics of the Transmission Control Protocol/Internet Protocol (TCP/IP) over Frame Relay/Asynchronous Transfer Mode (FR/ATM) Service Interworking. The conclusions are based on empirical TCP/IP performance test results collected on a FR/ATM Service Interworking testbed, architected with commercially available IP, FR, and ATM equipment. © 1998 John Wiley & Sons, Ltd.     

TCP/IP enhancements for satellite networks

With the emerging market for high-mobility remote access broadband services, satellite networks are becoming increasingly popular. Although the ubiquitous TCP/IP protocol is widely used to provide reliable data delivery in terrestrial networks, it faces many challenges in satellite environments. These stem from the inherent features of satellite channels, such as large delays, increased error rates, and bandwidth asymmetry. To address these concerns, a variety of solutions have been proposed. These include direct TCP enhancements to better tune the TCP/IP stack and additional selective acknowledgment mechanisms. Other proposals use advanced interworking to reduce sensitivity to various channel features. Examples include multiple TCP sessions, link-layer interworking, and ACK control schemes. Improving onboard satellite features (buffer management, flow control) can also provide benefits for TCP/IP transport. The objective of this article is to present the various solutions and discuss their possible trade-offs. Overall, there exists a rich set of alternatives to meet the challenges in this important arena     

Transport of IP controlled-load service over ATM networks

A hot topic in the research community is the interworking of integrated IP services, defined by the IETF, with ATM, a technology designed for the transport of multimedia traffic. In particular, the IETF recommends the ATM CBR, nrt-VBR, and ABR service classes for the transport of controlled-load service (CLS) over ATM. In this work the first two alternatives are examined and compared, in order to determine which is the most convenient for CLS. The comparison is carried out in terms of QoS and amount of buffering required at the IP/ATM routers. These performance parameters strongly depend on the parameters of the CLS token bucket shaper, that is, the token bucket depth and token bucket input buffer capacity. A nrt-VBR connection is able to offer a given performance level over a range of values of the token bucket parameters wider than that permitted by a CBR connection. Moreover, in all the examined cases the nrt-VBR connection needs a relatively small ATM buffer at the IP/ATM routers. Since, in order to guarantee acceptable performance, some CLS requests must be rejected, the utilization of the ATM link can be relatively low; this drawback can be overcome by serving the rejected CLS requests as classic best-effort traffic     

Cross-layer optimization in TCP/IP networks

TCP-AQM can be interpreted as distributed primal-dual algorithms to maximize aggregate utility over source rates. We show that an equilibrium of TCP/IP, if exists, maximizes aggregate utility over both source rates and routes, provided congestion prices are used as link costs. An equilibrium exists if and only if this utility maximization problem and its Lagrangian dual have no duality gap. In this case, TCP/IP incurs no penalty in not splitting traffic across multiple paths. Such an equilibrium, however, can be unstable. It can be stabilized by adding a static component to link cost, but at the expense of a reduced utility in equilibrium. If link capacities are optimally provisioned, however, pure static routing, which is necessarily stable, is sufficient to maximize utility. Moreover single-path routing again achieves the same utility as multipath routing at optimality.     

DRED: a random early detection algorithm for TCP/IP networks

It is now widely accepted that a RED [2] controlled queue certainly performs better than a drop‐tail queue. But an inherent weakness of RED is that its equilibrium queue length cannot be maintained at a preset value independent of the number of TCP active connections. In addition, RED's optimal parameter setting is largely correlated with the number of connections, the round‐trip time, the buffer space, etc. In light of these observations, we propose DRED, a novel algorithm which uses the basic ideas of feedback control to randomly discard packets with a load‐dependent probability when a buffer in a router gets congested. Over a wide range of load levels, DRED is able to stabilize a router queue occupancy at a level independent of the number of active TCP connections. The benefits of stabilized queues in a network are high resources utilization, predictable maximum delays, more certain buffer provisioning, and traffic‐load‐independent network performance in terms of traffic intensity and number of connections. Copyright © 2002 John Wiley & Sons, Ltd.     

The available bit rate service for data in ATM networks

The authors describe the purpose of the available bit rate (ABR) service in the context of other ATM services and review the agreements on the traffic control mechanism. They describe the recent progress towards standardization of the new service. The general nature of the ABR service is described in comparison with constant bit rate (CBR) and variable bit rate (VBR) services     

TCP／IP协议对ATM网的适配

当前在Ｉｎｔｅｒｎｅｔ和其他计算机网上广泛使用的ＴＣＰ／ＩＰ协议与未来准备在Ｂ－ＩＳＤＮ上应用的ＡＴＭ技术之间的有机结合，是一个战略性的技术课题。本文以ＴＣＰ／ＩＰ协议的体系结构及其作用、ＡＴＭ信号元结构与ＡＴＰ适应层为依据，对ＩＰ与ＡＴＭ适配的种种构想（其中包括协议复用、封装、地址分辨和连接管理等项功能的实现）做了轮廓性的探讨。     

A congestion control framework for available bit rate service in ATM networks

Fair packet discarding (FPD) is a mechanism which provides incentives to users to co‐operate in congestion control, so that ATM networks can operate more efficiently. In this paper, we propose a simple congestion control framework consisting of a combination of FPD and ATM Forum's explicit‐rate flow control scheme. We show by simulations that this frame‐work achieves all objectives of congestion control: (1) control congestion, (2) achieve high user‐perceived throughput, (3) maintain fairness and (4) provide protection from misbehaving users. In addition, we equip this framework with two enhancements to improve efficiency, and reduce end‐to‐end delay. Copyright © 2002 John Wiley & Sons, Ltd.     

TCP/IP与ATM的互联及网关功能的实现

人们在工作、学习和生活中提出了更多的通信业务,如高清晰度电视、电视会议、可视电话、点播电视等。对这些多功能化的需求,现今的任何网(包括电话网、用户电报网、公用分组交换数据网、电路交换数据网、数字数据网、LAN、WAN、MAN、ISDN、CATV等)都无法完成。因此,无论从用户角度还是从网络经营者的角度,都希望建立一个单一的网络。这个网络便是B-ISDN(宽带综合业务数字网)。文中主要讨论TCP/IP与ATM的互联,以及网关如何实现TCP/IP与ATM(异步传输模式)之间的协议转换。简单介绍了ATM的概念及分层模型,给出了IP与ATM结合的多种解决方式。     

A simple mechanism for stabilizing network queues in TCP/IP networks

In this paper we determine the stability bounds for the DRED active queue management (AQM) algorithm using a previously developed nonlinear dynamic model of TCP. We develop a second‐order linear model with time delay by linearizing the nonlinear model. Using the Pade approximation of time‐delayed system e^?R0s, where R₀ is the delay in the system, we then determine the range of stabilizing gains of DRED when controlling the second‐order system with time delay R₀. We also present examples showing the stability bounds of the DRED controller gain for networks with different parameters such as link capacity, load level, and round‐trip time. In addition, we describe an efficient implementation of the DRED AQM algorithm. Copyright © 2006 John Wiley & Sons, Ltd.     

Low‐loss TCP/IP header compression for wireless networks

Wireless is becoming a popular way to connect mobile computers to the Internet and other networks. The bandwidth of wireless links will probably always be limited due to properties of the physical medium and regulatory limits on the use of frequencies for radio communication. Therefore, it is necessary for network protocols to utilize the available bandwidth efficiently. Headers of IP packets are growing and the bandwidth required for transmitting headers is increasing. With the coming of IPv6 the address size increases from 4 to 16 bytes and the basic IP header increases from 20 to 40 bytes. Moreover, most mobility schemes tunnel packets addressed to mobile hosts by adding an extra IP header or extra routing information, typically increasing the size of TCP/IPv4 headers to 60 bytes and TCP/IPv6 headers to 100 bytes. In this paper, we provide new header compression schemes for UDP/IP and TCP/IP protocols. We show how to reduce the size of UDP/IP headers by an order of magnitude, down to four to five bytes. Our method works over simplex links, lossy links, multi‐access links, and supports multicast communication. We also show how to generalize the most commonly used method for header compression for TCP/IPv4, developed by Jacobson, to IPv6 and multiple IP headers. The resulting scheme unfortunately reduces TCP throughput over lossy links due to unfavorable interaction with TCP's congestion control mechanisms. However, by adding two simple mechanisms the potential gain from header compression can be realized over lossy wireless networks as well as point‐to‐point modem links. This revised version was published online in June 2006 with corrections to the Cover Date.     

Network management of TCP/IP networks: present and future

Recent network management activities in the TCP/IP community have focused on standardizing two network management protocols-Simple Network Management Protocol (SNMP) and Common Management Information Services and Protocol Over TCP/IP (CMOT)-that provide for the exchange of management information. The current SNMP and CMOT approaches to TCP/IP network management are compared from several different perspectives; comparisons are based on both theory and knowledge gained from actual implementation experiences. The current level of user and vendor acceptance for these two protocols is examined and explained, and ongoing standardization efforts are summarized. Relevant ongoing work is summarized, and trends over the next few years are discussed     

EVA: a better TCP version for resource‐insufficient networks

TCP Vegas exhibits unfair congestion avoidance mechanism, which aggravates when there are insufficient network resources to accommodate buffer space of a pipe (bandwidth delay product). To remedy this shortcoming, we propose an Enhanced VegAs (EVA) that employs three auxiliary mechanisms: Δ revision, congestion detection and congestion tendency detection. A 2^k factorial design with replications is used to study the effect of the three mechanisms. Our results show that TCP EVA achieves better performance than Vegas under various network conditions. Furthermore, congestion avoidance schemes, such as TCP EVA, perform much better than congestion control schemes, such as TCP Reno, in resource‐insufficient networks. Copyright © 2002 John Wiley & Sons, Ltd.     

Connectionless service for public ATM networks

The interconnection of connectionless LANs and MANs over a connection-oriented B-ISDN presents a dilemma with regard to efficient interoperability. In order to provide solutions to this dilemma, there have been many techniques proposed. We describe and evaluate the architectural and conceptual alternatives available for the implementation of a connectionless service for public ATM networks. We also suggest a scheme through which best-effort connectionless service may be provided in public ATM networks using connectionless servers and hop-by-hop flow control. There is as yet no consensus on issues such as bandwidth management, address resolution, connectionless server forwarding modes, or connectionless server topology, and therefore much of the area remains open for research. These research issues must be addressed, however, in order to realize the interoperability, and thereby the acceptance, of ATM     

区分服务网络中的TCP友好公平体系

区分服务(Differentiated Services)是IETF为实现IP服务质量(QoS)而定义的一个体系结构。研究表明,在该体系中存在不公平问题,该文将TCP友好(TCP Friendly)的概念引入到DS网络中,并定义了DS网络中的TCP友好的公平性,仿真验证了目前IETF定义的流量调节(TrafficConditioning)以及丢包策略等机制不能很好地实现TCP友好公平性,因而提出了直接拥塞控制机制来实现这一公平性。     

Design and evaluation of multicast GFR for supporting TCP/IP over hybrid wired/wireless ATM

The major challenges of designing multicast traffic control protocols for a combined wired/wireless network are the varying transmission characteristics (bandwidth, error, and propagation delay) of the wireless and wired media, and the different, possibly conflicting frame rate requests from multiple sources. To address these issues, in this paper we design and evaluate new unicast and multicast guaranteed frame rate (GFR) schemes for supporting TCP/IP traffic over a combined wired/wireless ATM network. We first propose a new, flexible weighted buffer management, and a frame‐based virtual spacing (VS) mechanism implementing weighted fair queueing. The unicast GFR scheme is based on the integration of the new weighted buffer management, and either cell‐based or frame‐based VS. It is then extended to support multicast GFR flows. The multicast scheme presented in this paper is the first multicast GFR scheme appeared in the literature. These schemes are carefully evaluated over several network configuration, supporting heterogeneous TCP/IP traffic with various frame rates. Simulation results show that the new schemes guarantee the minimum rates requested, provide excellent fairness, and achieve reasonably high efficiency. The new schemes may be extended to provide differentiated service in both IP and mobile IP frame work. This revised version was published online in July 2006 with corrections to the Cover Date.