This paper presents a tutorial on transport layer concepts and terminology, and a survey of transport layer services and protocols. The transport layer protocol TCP is used as a reference point, and compared and contrasted with nineteen other protocols designed over the past two decades. The service and protocol features of twelve of the most important protocols are summarized in both text and tables

Bandwidth efficiency in communication networks can require supporting traffic with greatly differing quality of service requirements, particularly in terms of delay. The facilities needed to permit users to take advantage of excess bandwidth include (i) a method of conveying network state information to the source, and (ii) a method for guaranteeing quality of service requirements. Reservation oriented rate allocating servers (RASs) provide QOS guarantees while reservationless RASs provide fairness. This paper gives a technique for modifying a reservation oriented RAS to accommodate reservationless traffic and allow users with bandwidth reservations to take advantage of excess network bandwidth. In addition, this paper shows how to use the packet pair bandwidth probing technique with these new RASs to provide effective and efficient congestion control. 1. Introduction Network efficiency may be enhanced by carrying traffic with widely different quality of service (QOS) requirements. Co...

A number of distributed corporate intranets can be connected together using the ABR service of an ATM network to form a virtual private network. However, because the available bandwidth of the ABR connection may be occasionally reduced by the ATM network, severe congestion may occur at the gateway connecting an intranet to the ATM network. Such congestion is likely to cause packet loss at the gateway which in turn may significantly reduce the throughput of intranet applications. In order to address the above problem, we propose a novel rate-based flow control mechanism, called Z. Host Gateway Rate Control Protocol HG-RCP , between the gateway and the intranet hosts. The control mechanism of HG-RCP is modeled as a classical feedback control system and the control parameters are selected to guarantee the stability of the system. The performance of HG-RCP is evaluated using simulation for both limited and unlimited buffers at the gateway. Simulation results confirm that HG-RCP is very stable and fair in allocating the ABR bandwidth among the competing sources in the intranet. It is also demonstrated via simulation that HG-RCP can significantly reduce the amount of buffers required at the gateway and improve the performance of TCPrIP applications in the intranets. q 1999 Elsevier Science B.V. All rights reserved.

In this paper we analyze a novel paradigm of reliable communication which is not based on the traditional timeout-and-retransmit mechanism of TCP. Our approach, which we call FBP (Fountain Based Protocol), consists of using a digital fountain encoding which guarantees that duplicate packets are almost impossible. By using Game Theory, we analyze the behavior of TCP and FBP in the presence of congestion. We show that hosts using TCP have an incentive to switch to an FBP approach, obtaining a higher goodput. Furthermore, we also show that a Nash equilibrium occurs when all hosts use FBP (i.e., when FBP hosts act in an absolutely selfish manner injecting packets into the network as fast as they can and without any kind of congestion control approach). At this equilibrium, the performance of the network is similar to the performance obtained when all hosts comply with TCP. Regarding the interaction of hosts using FBP at different rates, our results show that the Nash equilibrium is reached when all hosts send at the highest possible rate, and, as before, that the performance of the network in such a case is similar to the obtained when all hosts comply with TCP.

The stability of the current Internet depends on the end-to-end congestion control mechanism provided by TCP. Recently, popular multimedia applications (RealAudio, RLM) started using more aggressive forms of congestion control. It is not clear yet how wide uncooperation will become in Internet, but it seems dangerous to base the Internet congestion control solely on the assumption of end-host cooperation. In this paper we examine the impact of router based congestion control mechanisms in the presence of uncooperative trac. We consider two classes of congestion control mechanisms (1) fair queuing mechanisms (FRED, DRR, CSFQ) and (2) feedback based mechanisms (ERAF, NBP). Within the subclass of feedback based mechanisms we distinguish between binary feedback mechanisms (ERAF) and rate feedback mechanisms (NBP). The criteria we use for our comparison are: eciency, fairness, convergence and scalability. Our results indicate that the feedback based mechanisms are able to shield responsive trac competing with uncooperative trac on a single link with the same eciency as the fair queuing mechanisms. For more complex topologies, the feedback based mechanisms oer better bandwidth allocations than the fair queuing mechanisms alone. In particular, the binary feedback mechanism we examine is able to provide allocations close to proportional fairness.

The congestion control algorithm embedded in the 4.3-Tahoe BSD TCP implementation has dramatically improved congestion control over the Internet. However, several recent simulation studies on the dynamics of this algorithm has revealed that the algorithm exhibits clear oscillatory patterns in sending window size, round trip delay and bottleneck queue length. In this paper, we present a new congestion signal scheme and a dual traffic adjustment strategy. Simulation results show that our modifications can eliminate the periodic packet losses and substantially reduce the traffic oscillation. 1. INTRODUCTION The congestion control algorithm embedded in the 4.3-Tahoe BSD TCP implementation was developed by Van Jacobson when the Internet was experiencing congestion collapse[10]. The algorithm has dramatically improved congestion control over the Internet and is now regarded as an Internet Standard [8]. More recently there have been several simulation studies on the dynamics of this algorith...

draft-irtf-iccrg-cc-rfcs-02 By submitting this Internet-Draft, each author represents that any applicable patent or other IPR claims of which he or she is aware have been or will be disclosed, and any of which he or she becomes aware will be disclosed, in accordance with Section 6 of BCP 79. 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 draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at

This paper is a brief description of (i) -- (v) and the rationale behind them. (vi) is an algorithm recently developed by Phil Karn of Bell Communications Research, described in [15]. (vii) is described in a soon-to-be-published RFC (ARPANET "Request for Comments")

Current congestion control mechanisms in today's Internet rely on end-to-end TCP congestion avoidance algorithms that back off sources when congestion occurs, detected by packet loss or explicit congestion notification signals. But one of the major setbacks of such mechanisms is making sure that all sources respond correctly in applying congestion avoidance measures during periods of high congestion; non-compliant sources can aggravate the network congestion state and yield to unfairness, worsening the state of the network.

The challenges of optimizing end-to-end performance over diverse Internet paths has driven widespread adoption of inpath optimizers, which can destructively interfere with TCP’s end-to-end semantics and with each other, and are incompatible with end-to-end IPsec. We identify the architectural cause of these conflicts and resolve them in Tng, an experimental next-generation transport services architecture, by factoring congestion control from end-to-end semantic functions. Through a technique we call queue sharing, Tng enables in-path devices to interpose on, split, and optimize congestion controlled flows without affecting or seeing the end-to-end content riding these flows. Simulations show that Tng’s decoupling cleanly addresses several common performance problems, such as communication over lossy wireless links and reduction of buffering-induced latency on residential links. A working prototype and several incremental deployment paths suggest Tng’s practicality.