Queuing dynamics and single-link stability of delay-based window congestion control

Accurate modeling of queueing dynamics is important in the design and analysis of Internet congestion control. However, as demonstrated in this paper, existing window-based queueing models [26,30] are often not capable of precisely capturing the transient behavior (i.e., self-clocking and burstiness) of TCP-like protocols and their resulting analysis may be inaccurate in practice. As one example, we show that stability conditions of FAST TCP based on traditional queuing models [17] that do not considering transient dynamics of the queues are inconsistent with ns2 simulations. We explain the origin of this problem and overcome it by developing a novel approach called Self-clocking Queuing Model (SQM) that accurately describes both the steady-state and transient queuing behavior of window-based control systems. Using SQM and explicitly incorporating control interval h"i in the queuing model and derive a sufficient condition for its local stability under homogeneous delay, which strengthens prior results [29,30] obtained using traditional queuing models.

[1]  Fernando Paganini,et al.  FAST TCP: from theory to experiments , 2005, IEEE Network.

[2]  Mario Gerla,et al.  Modeling Window Based Congestion Control Protocols with Many Flows , 1999, Perform. Evaluation.

[3]  Larry L. Peterson,et al.  TCP Vegas: new techniques for congestion detection and avoidance , 1994 .

[4]  P. Glendinning Stability, Instability and Chaos: An Introduction to the Theory of Nonlinear Differential Equations , 1994 .

[5]  Jon Crowcroft,et al.  A new congestion control scheme: slow start and search (Tri-S) , 1991, CCRV.

[6]  Marco Ajmone Marsan,et al.  A detailed and accurate closed queueing network model of many interacting TCP flows , 2001, Proceedings IEEE INFOCOM 2001. Conference on Computer Communications. Twentieth Annual Joint Conference of the IEEE Computer and Communications Society (Cat. No.01CH37213).

[7]  Ren Wang,et al.  TCP with sender-side intelligence to handle dynamic, large, leaky pipes , 2005, IEEE Journal on Selected Areas in Communications.

[8]  Gregorio Procissi,et al.  Sender-Side TCP Modifications: Performance Analysis and Design Guidelines , 2004, Cluster Computing.

[9]  Saverio Mascolo,et al.  Performance evaluation and comparison of Westwood+, New Reno, and Vegas TCP congestion control , 2004, CCRV.

[10]  Jiantao Wang,et al.  Modelling and stability of FAST TCP , 2005, Proceedings IEEE 24th Annual Joint Conference of the IEEE Computer and Communications Societies..

[11]  Lachlan L. H. Andrew,et al.  ACK-Clocking Dynamics: Modelling the Interaction between Windows and the Network , 2008, IEEE INFOCOM 2008 - The 27th Conference on Computer Communications.

[12]  Stability , 1973 .

[13]  Cheng Jin,et al.  FAST TCP: Motivation, Architecture, Algorithms, Performance , 2006, IEEE/ACM Transactions on Networking.

[14]  Sally Floyd,et al.  HighSpeed TCP for Large Congestion Windows , 2003, RFC.

[15]  Ren Wang,et al.  TCP westwood: Bandwidth estimation for enhanced transport over wireless links , 2001, MobiCom '01.

[16]  Lachlan L. H. Andrew,et al.  An Improved Link Model for Window Flow Control and Its Application to FAST TCP , 2009, IEEE Transactions on Automatic Control.

[17]  Richard G. Baraniuk,et al.  TCP-Africa: an adaptive and fair rapid increase rule for scalable TCP , 2005, Proceedings IEEE 24th Annual Joint Conference of the IEEE Computer and Communications Societies..

[18]  Steven H. Low,et al.  Stabilized Vegas , 2003, IEEE INFOCOM 2003. Twenty-second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No.03CH37428).

[19]  Vishal Misra,et al.  Fluid-based analysis of a network of AQM routers supporting TCP flows with an application to RED , 2000, SIGCOMM.

[20]  Amr Awadallah,et al.  TCP-BFA: Buffer Fill Avoidance , 1998, HPN.

[21]  Jon Crowcroft,et al.  Eliminating periodic packet losses in the 4.3-Tahoe BSD TCP congestion control algorithm , 1992, CCRV.

[22]  Dmitri Loguinov,et al.  Emulating AQM from end hosts , 2007, SIGCOMM.

[23]  Jiantao Wang,et al.  Local stability of FAST TCP , 2004, 2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601).

[24]  Injong Rhee,et al.  Binary increase congestion control (BIC) for fast long-distance networks , 2004, IEEE INFOCOM 2004.

[25]  Tom Kelly,et al.  Scalable TCP: improving performance in highspeed wide area networks , 2003, CCRV.

[26]  K. Jacobsson,et al.  ACK-clock Dynamics in Network Congestion Control - An Inner Feedback Loop with Implications on Inelastic Flow Impact , 2006, Proceedings of the 45th IEEE Conference on Decision and Control.

[27]  Vishal Misra,et al.  Fluid-based analysis of a network of AQM routers supporting TCP flows with an application to RED , 2000, SIGCOMM 2000.

[28]  Ramesh Johari,et al.  End-to-end congestion control for the internet: delays and stability , 2001, TNET.

[29]  Donald F. Towsley,et al.  Analysis and design of controllers for AQM routers supporting TCP flows , 2002, IEEE Trans. Autom. Control..

[30]  Mark Handley,et al.  Congestion control for high bandwidth-delay product networks , 2002, SIGCOMM '02.

[31]  Ao Tang,et al.  Linear stability analysis of FAST TCP using a new accurate link model , 2006 .

[32]  Marco Ajmone Marsan,et al.  Modeling short-lived TCP connections with open multiclass queuing networks , 2004, Comput. Networks.

[33]  MarsanMarco Ajmone,et al.  Closed queueing network models of interacting long-lived TCP flows , 2004 .

[34]  Xuezheng Liu,et al.  D3S: Debugging Deployed Distributed Systems , 2008, NSDI.

[35]  Dmitri Loguinov,et al.  JetMax: Scalable Max-Min Congestion Control for High-Speed Heterogeneous Networks , 2006, Proceedings IEEE INFOCOM 2006. 25TH IEEE International Conference on Computer Communications.

[36]  Lakshminarayanan Subramanian,et al.  One more bit is enough , 2005, SIGCOMM '05.