Understanding TCP dynamics in a di erentiated services internet

A large number of Internet applications are sensitive to overload conditions in the network. While these applications have been designed to adapt somewhat to the varying conditions in the Internet, they can beneet greatly from an increased level of predictability in network services. We propose minor extensions to the packet queueing and discard mechanisms used in routers, coupled with simple control mechanisms at the source that enable the network to guarantee minimal levels of throughput to diierent sessions while sharing the residual network capacity in a cooperative manner. The service realized by the proposed mechanisms is an interpretation of the controlled-load service being standardized by the ietf. Although controlled-load service can be used in conjunction with any transport protocol, our focus in this paper is on understanding its interaction with tcp. Speciically, we study the dynamics of tcp traac in an integrated services network that simultaneously supports both best-eeort and controlled-load sessions. In light of this study, we propose and experiment with several variations to tcp's control mechanisms with the objective of ne-tuning them for an integrated services environment. We then investigate the eeect of network transients, such as changes in traac load and in service levels, on the performance of controlled-load as well as best-eeort connections. To capture the evolution of integrated services in the Internet, we also consider situations where only a selective set of routers are capable of providing service diierentiation between best-eeort and controlled-load traac. Finally, we show how the service mechanisms proposed here can be embedded within more elaborate packet and link scheduling frameworks in a fully-evolved integrated services Internet.

[1]  V. Jacobson,et al.  Congestion avoidance and control , 1988, SIGCOMM '88.

[2]  S. Shenker,et al.  Observations on the dynamics of a congestion control algorithm: the effects of two-way traffic , 1991, SIGCOMM '91.

[3]  Erwin P. Rathgeb Modeling and Performance Comparison of Policing Mechanisms for ATM Networks , 1991, IEEE J. Sel. Areas Commun..

[4]  QUTdN QeO,et al.  Random early detection gateways for congestion avoidance , 1993, TNET.

[5]  S. Jamaloddin Golestani,et al.  A self-clocked fair queueing scheme for broadband applications , 1994, Proceedings of INFOCOM '94 Conference on Computer Communications.

[6]  Scott Shenker,et al.  Integrated Services in the Internet Architecture : an Overview Status of this Memo , 1994 .

[7]  Van Jacobson,et al.  Link-sharing and resource management models for packet networks , 1995, TNET.

[8]  George Varghese,et al.  Redesigning the BSD Callout and Timer Facilities , 1995 .

[9]  Eecient Fair Queuing Using Deecit round Robin , 1995 .

[10]  Peter B. Danzig,et al.  A measurement-based admission control algorithm for integrated services packet networks , 1995, SIGCOMM '95.

[11]  Partho Pratim Mishra Effect of leaky bucket policing on TCP over ATM performance , 1996, Proceedings of ICC/SUPERCOMM '96 - International Conference on Communications.

[12]  Albert G. Greenberg,et al.  Hardware-eecient Fair Queueing Architectures for High-speed Networks , 1996 .

[13]  Janey C. Hoe,et al.  Improving the start-up behavior of a congestion control scheme for TCP , 1996, SIGCOMM '96.

[14]  H. Vin,et al.  Start-time fair queueing: a scheduling algorithm for integrated services packet switching networks , 1996, SIGCOMM '96.

[15]  Raj Jain,et al.  UBR+: improving performance of TCP over ATM-UBR service , 1997, Proceedings of ICC'97 - International Conference on Communications.