Sally Floyd

University of California, Berkeley | UCB

RFC 5681 documents the following four intertwined TCP congestion control algorithms: slow start, congestion avoidance, fast retransmit, and fast recovery. RFC 5681 explicitly allows certain modifications of these algorithms, including modifications that use the TCP Selective Acknowledgment (SACK) option (RFC 2883), and modifications that respond to...

Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. Internet-Drafts are d...

This document describes a possible congestion control mechanism for acknowledgement (ACKs) traffic in TCP. The document specifies an end-to-end acknowledgement congestion control mechanism for TCP that uses participation from both TCP hosts: the TCP data sender and the TCP data receiver. The TCP data sender detects lost or Explicit Congestion Notif...

TCP-Friendly Rate Control (TFRC) is a congestion control mechanism for unicast flows operating in a best-effort Internet environment (RFC3448). This document introduces Faster Restart, an optional mechanism for safely improving the behavior of interacti ve flows that use TFRC. Faster Restart is proposed for use with both the default TFRC and with t...

This document presents some observations on "simple best-effort traffic", defined loosely for the purposes of this document as Internet traffic that is not covered by Quality of Service (QOS) mechanisms, congestion-based pricing, cost-based fairness, admissions control, or the like. One observation is that simple best-effort traffic serves a useful...

This document describes extensions for the flow control of the Transmission Control Protocol (TCP) that avoid interactions with fast startup congestion control mechanisms, in particular the Quick-Start TCP extension. Quick-Start is an optional TCP extension that allows to start data transfers with a large congestion window, using feedback of the ro...

This document describes tools for the e valuation of simulation and testbed scenarios used in research on Internet congestion control mechanisms. We believe that research in congestion control mechanisms has been seriously hampered by the lack of good models underpinning analysis, simulation, and testbed experiments, and that tools for the evaluati...

URL: www.icir.org/tbit/

14 pages This document describes methods to avoid interactions between the flow control of the Transmission Control Protocol (TCP) and the Quick-Start TCP extension. Quick-Start is an optional TCP congestion control mechanism that allows hosts to determine an allowed sending rate from feedback of routers along the path. With Quick-Start, data trans...

Determining an appropriate sending rate when beginning data transmission into a network with unknown characteristics is a fundamental issue in best-effort networks. Traditionally, the slow-start algorithm has been used to probe the network path for an appropriate sending rate. This paper provides an initial exploration of the efficacy of an alterna...

There have been a number of proposals for alternate semantics for the Explicit Congestion Notification (ECN) field in the IP header RFC 3168. This document discusses some of the issues in defining alternate semantics for the ECN field, and specifies requirements for a safe coexistence in an Internet that could include routers that do not understand...

DCCP, the Datagram Congestion Control Protocol, is a new transport protocol in the TCP/UDP family that provides a congestion-controlled flow of unreliable datagrams. Delay-sensitive applications, such as streaming media and telephony, prefer timeliness to reliability. These applications have historically used UDP and implemented their own congestio...

This document contains the profile for Congestion Control Identifier 2, TCP-like Congestion Control, in the Datagram Congestion Control Protocol (DCCP) (DCCP). DCCP implements a congestion-controlled, unreliable flo wo fd atagrams suitable for use by applications such as streaming media. The TCP-lik eC ongestion Control CCID is used by senders who...

This document contains the profile for Congestion Control Identifier 3, TCP-Friendly Rate Control (TFRC), in the Datagram Congestion Control Protocol (DCCP). CCID 3 should be used by senders that want a TCP-friendly sending rate, possibly with Explicit Congestion Notification (ECN), while minimizing abrupt rate changes.

This document gives the problem statement underlying the development of an unreliable transport protocol incorporating end-to-end congestion control. This is also the problem statement underlying the development of DCP, the Datagram Control Protocol. DCP implements a congestion-controlled, unreliable flow of datagrams suitable for use by applicatio...

This paper reports on the project plan to develop a new major version of the popular ns-2 networking simulator. The authors have organized an NSF-funded, four-year community infrastructure project to develop the next version of ns. The project will also be oriented towards community development and open source software practices to encourage partic...

Viewgraphs, talk at Hamilton Institute, Ireland

How might the computing and communications world be materially different in 10 to 15 years, and how might we define a research agenda that would get us to that world?

In this paper we explore the evolution of both the Internet's most heavily used transport protocol, TCP, and the current network environment with respect to how the network's evolution ultimately impacts end-to-end protocols. The traditional end-to-end assumptions about the Internet are increasingly challenged by the introduction of intermediary ne...

This paper explores the Quick-Start mechanism, designed to allow transport protocols to explicitly request permis-sion from the routers along a network path to send at a higher rate than normally allowed by traditional congestion control mechanisms. If the routers are underutilized, they may approve the sender's request for a higher sending rate; o...

Wireless links have intrinsic characteristics that affect the performance of transport protocols; these include variable bandwidth, corruption, channel allocation delays, and asymmetry. In this paper we review simulation models for cellular, WLAN and satellite links used in the design of transport protocols, and consider the interplay between wirel...

Creating an experimental infrastructure for developing next-generation information security technologies.

This note proposes a modification for TCP's slow-start for use with TCP connections with large congestion windows. For TCP connections that are able to use congestion windows of thousands (or tens of thousands) of MSS-sized segments (for MSS the sender's MAXIMUM SEGMENT SIZE), the current slow-start procedure can result in increasing the congestion...

In the absence of support for Selective Acknowledgments (SACK) from its peer, TCP suffers from ambiguity regarding Duplicate Acknowledgments after a Retransmit Timeout. In particular, TCP is not able to determine if Duplicate Acknowledgments result from packets that were retransmitted unnecessarily, or from an earlier packet that was lost. Conseque...

In this paper we explore the evolution of both the Internet's most heavily used transport protocol, TCP, and the current network environment with respect to how the network's evolution ultimately impacts end-to-end protocols. The traditional end-to-end assumptions about the Internet are increasingly challenged by the introduction of intermediary ne...

This document proposes HighSpeed TCP, a modification to TCP's congestion control mechanism for use with TCP connections with large congestion windows. The congestion control mechanisms of the current Standard TCP constrains the congestion windows that can be achieved by TCP in realistic environments. For example, for a Standard TCP connection...

TCP performs poorly on paths that reorder packets significantly, where it misinterprets out-of-order delivery as packet loss. The sender responds with a fast retransmit though no actual loss has occurred. These repeated false fast retransmits keep the sender's window small, and severely degrade the throughput it attains. Requiring nearly in-order d...

In this paper we consider the potential for congestion collapse in a range of network scenarios. In particular, we are interested in the effect of the topology, the scheduling discipline (FIFO or FQ scheduling), the level of statistical multiplexing, the traffic characteristics, and other factors. We consider topologies more complex than a single c...

We provide a short overview of Datagram Congestion Control Protocol (DCCP), which implements a congestion-controlled, unreliable flow of datagrams suitable for use by applications such as streaming media. (This document consists mostly of excerpts from (DCCP). Section 1 serves as a brief introduction to DCCP ,while the remaining sections serve as...

This document specifies TCP-Friendly Rate Control (TFRC). TFRC is a congestion control mechanism for unicast flows operating in a besteffort Internet environment. It is reasonably fair when competing for bandwidth with TCP flows, but has a much lower variation of throughput over time compared with TCP, making it more suitable for applications such...

FIFO queueing is simple but does not protect traffic from high-bandwidth flows, which include not only flows that fail to use end-to-end congestion control, but also short round-trip time TCP flows. At the other extreme, per-flow scheduling mechanisms provide max-min fairness but are more complex, keeping state for all flows going through the route...

Networking researchers work from mental models of the Inter- net's important properties. The scenarios used in simulations and experiments reveal aspects of these mental models (including our own), often including one or more of the following implicit as- sumptions: Flows live for a long time and transfer a lot of data. Simple topologies, like a "d...

This document specifies an increase in the permitted initial window for Transmission Control Protocol (TCP) from one segment to roughly 4K bytes. This document discusses the advantages and disadvantages of such a change, outlining experimental results that indicate the costs and benefits of such a change to TCP.

An abstract is not available.

TCP performs poorly on paths that reorder packets significantly, where it misinterprets out-of-order delivery as packet loss. The sender responds with a fast retransmit though no actual loss has occurred. These repeated false fast retransmits keep the sender's window small, and severely degrade the throughput it attains. Persistent reordering occas...

Abstract The current Internet infrastructure has very few built-in p rotection mechanisms,and is therefore quite vulnerable to attacks and failures. In particular, recent events have illus trated the Internet’s vulnerability to both Distributed Denial of Service attacks and “flash crowds” in which one or more links in the network (or servers at the...

This document describes a framework for the standardization of bulkdata reliable multicast transport. It builds upon the experience gained during the deployment of several classes of contemporary reliable multicast transport, and attempts to pull out the commonalities between these classes of protocols into a number of building blocks. To that end,...

This document suggests general architectural and policy questions tobe addressed in our work in the IETF. We note that this documentcontains questions to be addressed, as opposed to guidelines orarchitectural principles to be followed.

FIFO queueing is simple but does not protect traffic from high-bandwidth flows, which include not only flows that fail to use end-to-end congestion control, but also short round-trip time TCP flows. At the other extreme, per-flow scheduling mechanisms provide max-min fairness but are more complex, keeping state for all flows going through the route...

Viewgraphs, IEEE Annual Computer Communications Workshop

The RED active queue management algorithm allows network operators to simultaneously achieve high throughput and low average delay. However, the resulting average queue length is quite sensitive to the level of congestion and to the RED parameter settings, and is therefore not predictable in advance. Delay being a major component of the quality of...

Simulating how the global Internet behaves is an immensely challenging undertaking because of the network's great heterogeneity and rapid change. The heterogeneity ranges from the individual links that carry the network's traffic, to the protocols that interoperate over the links, the “mix” of different applications used at a site, and the levels o...

The recently developed notion of TCP-compatibility has led to a number of proposals for alternative congestion control algorithms whose long-term throughput as a function of a steady-state loss rate is similar to that of TCP. Motivated by the needs of some streaming and multicast applications, these algorithms seem poised to take the current TCP-do...

Most of the traffic in today's Internet is controlled by the Transmission Control Protocol (TCP). Hence, the performance of TCP has a significant impact on the performance of the overall Internet. TCP is a complex protocol with many user-configurable parameters and a range of different implementations. In addition, research continues to produce new...

This document specifies the incorporation of ECN (Explicit Congestion Notification) to TCP and IP, including ECN's use of two bits in the IP header. We begin by describing TCP's use of packet drops as an indication of congestion. Next we explain that with the addition of active queue management (e.g., RED) to the Internet infrastructure, where rout...

The current Internet infrastructure has very few built-in protection mechanisms and is therefore vulnerable to at- tacks and failures. In particular, recent events have il- lustrated the Internet' s vulnerability to both denial of service (DoS) attacks and flash crowds in which one or more links in the network (or servers at the edge of the network...

Abstract

This article discusses several changes to TCP's congestion control, either proposed or in progress. The changes to TCP include a limited transmit mechanism for transmitting new packets upon receipt of one or two duplicate acknowledgments, and a SACK-based mechanism for detecting and responding to unnecessary fast retransmits or retransmit timeouts....

Most of the traffic in today's Internet is controlled by the Transmission Control Protocol (TCP). Hence, the performance of TCP has a significant impact on the performance of the overall Internet. Since web traffic forms the majority of the TCP traffic, TCP implementations in today's web servers are of particular interest. However, TCP is a complex...

This document describes the messages and procedures of the Negative-acknowledgment (NACK) Oriented Reliable Multicast (NORM) protocol. This protocol is designed to provide end-to-end reliable transport of bulk data objects or streams over generic IP multicast routing and forwarding services. NORM uses a selective, negative acknowledgment mechanism...

This paper considers AIMD-based (Additive-Increase Multiplicative-Decrease) congestion control mechanisms that are TCP-compatible (i.e., that compete reasonably fairly with TCP), but that reduce their sending rate less sharply than does TCP in response to a single packet drop. The paper then briefly compares these smoother AIMD-based congestion con...

This paper proposes a mechanism for equation-based congestion control for unicast traffic. Most best-effort traffic in the current Internet is well-served by the dominant transport protocol TCP. However, traffic such as best-effort unicast streaming multimedia could find use for a TCP-friendly congestion control mechanism that refrains from reducin...

ADVANTAGES OF A COMMON SIMULATOR Multiprotocol network simulators provide a rich opportunity for efficient experimentation. Disparate research efforts using a common simulation environment can yield substantial benefits, including . improved validation of the behavior of existing protocols, . a rich infrastructure for developing new protocols, . th...

TCP's congestion window controls the number of packets a TCP flow may have in the network at any time. However, long periods when the sender is idle or application-limited can lead to the invalidation of the congestion window, in that the congestion window no longer reflects current information about the state of the network. In this paper we propo...

This paper proposes a mechanism for equation-based congestion control for unicast traffic. Most best-effort traffic in the current Internet is well-served by the dominant transport protocol, TCP. However, traffic such as best-effort unicast streaming multimedia could find use for a TCP-friendly congestion control mechanism that refrains from reduci...

This paper proposes a mechanism for equation-based congestion control for unicast traffic. Most besteffort traffic in the current Internet is well-served by the dominant transport protocol TCP. However, traffic such as best-effort unicast streaming multimedia could find use for a TCP-friendly congestion control mechanism that refrains from reducing...

Soft state protocols use periodic refresh messages to keep network state alive while adapting to changing network conditions; this has raised concerns regarding the scalability of protocols that use the soft-state approach. In existing soft state protocols, the values of the timers that control the sending of these messages, and the timers for agin...

This paper appeared in IEEE/ACM Transactions on Networking, Vol. 3 No. 4, August 1995 --- This paper discusses the use of link-sharing mechanisms in packet networks and presents algorithms for hierarchical link-sharing. Hierarchical link-sharing allows multiple agencies, protocol families, or traffic types to share the bandwidth on a link in a cont...

This paper considers the potentially negative impacts of an increasing deployment of non-congestion-controlled best-effort traffic on the Internet. These negative impacts range from extreme unfairness against competing TCP traffic to the potential for congestion collapse. To promote the inclusion of end-to-end congestion control in the design of fu...

RFC 2001 [RFC2001] documents the following four intertwined TCP congestion control algorithms: Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery. RFC 2581 [RFC2581] explicitly allows certain modifications of these algorithms, including modifications that use the TCP Selective Acknowledgement (SACK) option [MMFR96], and modificati...

This short note is written as a submission to the DARPA Workshop on Validation of Large Scale Network Simulation Models. The purpose of the workshop is to identify issues critical to the validation of network models, and also to develop potential strategies for the research community to address these issues. 1 Current network simulation models or t...

This document presents a strawman specification for TCP-Friendly Reliable Multicast Congestion Control (TFMCC). It is not a standard of any form. We do not suggest basing products on this specification. Eventually, if further research shows this specification to perform acceptably, we envisage moving this document into the IETF for standardization,...

An adaptive, highly scalable, and robust web caching system is needed to effectively handle the exponential growth and extreme dynamic environment of the World Wide Web. Our work presented last year sketched out the basic design of such a system [31]. This sequel paper reports our progress over the past year. To assist caches making Web query forwa...

Introduction In recent years, the Internet has grown signicantly in size and scope, and as a result new protocols and algorithms are being developed to meet changing operational requirements in the Internet. Examples of such requirements include quality of service support, multicast transport, security, mobile networking, and policy management. Dev...

This document illustrates the validation test for ECN (Explicit Congestion Notification) in the NS simulator [NS95]. Figure 1 shows a single simulation with two-way traffic, showing the use of the CE (Congestion Experienced) bit in the IP packet header, and the CWR (Congestion Window Reduction) and ECN-Echo bits in the TCP header. We refer to a pac...

URL: www.icir.org/floyd/tcp_friendly.html

An adaptive, highly scalable, and robust web caching system is needed to effectively handle the exponential growth and extreme dynamic environment of the World Wide Web. Our work presented last year sketched out the basic design of such a system. This sequel paper reports our progress over the past year. To assist caches making web query forwarding...

Simulation is an important tool in network protocol development, providing an effective way to perform controlled experiments, consider alternative designs, understand protocol interactions, and examine scales and topologies that are difficult to create in the laboratory. However, the scale and heterogeneity of today's networks create challenges fo...

this paper we call them cache groups.

End-to-end protocols measure network characteristics and react based on their estimates of network performance. Network dynamics can alter the topology significantly, and thereby affect protocol operation. Topology changes may result in routing pathologies (such as route loops, packet interleaving) , changes to the end-to-end path characteristics,...

This memo presents two recommendations to the Internet community concerning measures to improve and preserve Internet performance. It presents a strong recommendation for testing, standardization, and widespread deployment of active queue management in routers, to improve the performance of today's Internet. It also urges a concerted effort of rese...

This paper outlines efficient mechanisms to estimate the arrival rate of high-bandwidth flows for a router implementing RED active queue management. For such a router, the RED packet drop history constitutes a random sampling of the arriving packets; a flow with a significant fraction of the dropped packets is likely to have a correspondinglysignif...

Not Available

This paper describes scalable reliable multicast (SRM), a reliable multicast framework for light-weight sessions and application level framing. The algorithms of this framework are efficient, robust, and scale well to both very large networks and very large sessions. The SRM framework has been prototyped in wb, a distributed whiteboard application,...

Simulating how the global Internet data network behaves is an immensely challenging undertaking because of the net- work's great heterogeneity and rapid change. The hetero- geneity ranges from the individual links that carry the net- work's traffic, to the protocols that interoperate over the links, to the “mix” of different applications used at a...

This document is very much in progress. 1 Introduction

Multi-party applications are of great importance, and perhaps the greatest chal lenge is achieving both a) robustness in terms of adaptation to dynamic topology and group membership, and b) scalability in terms of bandwidth, state, and processing, as the size of a group grows. Scalabl e Reliable Multicast (SRM) (1) is a rich example of a robust des...

This note shows some of the tests that we use to verify that our simulator is performing the way that we intend it to perform.

this document are in the format used by our old simulator, tcpsim. All of these tests can be run in our new simulator ns with the command "test-all-red", and the input files are available in "test-suite-red.tcl". On each page, the graph shows the results of the simulation. For each graph, the x-axis shows the time in seconds. The y-axis shows the p...

this document consistent with the current CBQ code base in ns version 1. He has also i