FluNet: A hybrid internet simulator for fast queue regimes

Motivated by the scale and complexity of simulating large-scale networks, recent research has focused on hybrid fluid/packet simulators, where fluid models are combined with packet models in order to reduce simulation complexity as well as to track dynamics of end-sources accurately. However, these simulators still need to track the queuing dynamics of network routers, leading to considerable simulation complexity in a large-scale network model. In this paper, we propose a new hybrid simulator - FluNet - where queueing dynamics are not tracked, but instead, an equivalent rate-based model is used. The FluNet simulator is predicated on a fast-queueing regime at bottleneck routers, where the queue length fluctuates on a faster time-scale than end systems. This allows us to simulate large-scale systems, where the simulation ''time step-size'' is governed only by the time-scale of the end-systems, and not by that of the intermediate routers; whereas a queue-tracking based fluid simulator would require decreasingly smaller step-sizes as the system scale size increases. We validate our model using a ns-2 based implementation. Our results indicate a good match between packet systems and the associated FluNet system.

[1]  George Kesidis,et al.  Feasibility of fluid event-driven simulation for ATM networks , 1996, Proceedings of GLOBECOM'96. 1996 IEEE Global Telecommunications Conference.

[2]  Hao Jiang,et al.  Why is the internet traffic bursty in short time scales? , 2005, SIGMETRICS '05.

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

[4]  Donald F. Towsley,et al.  Inferring TCP connection characteristics through passive measurements , 2004, IEEE INFOCOM 2004.

[5]  Frank Kelly,et al.  Models for a self–managed Internet , 2000, Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.

[6]  Eddie Kohler,et al.  Internet research needs better models , 2003, CCRV.

[7]  Benjamin Melamed,et al.  Hybrid discrete-continuous fluid-flow simulation , 2001, SPIE ITCom.

[8]  R. Srikant,et al.  Rate-based versus queue-based models of congestion control , 2004, IEEE Transactions on Automatic Control.

[9]  Tim Roughgarden,et al.  Part III: routers with very small buffers , 2005, CCRV.

[10]  Mineo Takai,et al.  Integration of fluid-based analytical model with packet-level simulation for analysis of computer networks , 2001, SPIE ITCom.

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

[12]  Donald F. Towsley,et al.  On the efficiency of fluid simulation of networks , 2006, Comput. Networks.

[13]  R. Srikant,et al.  Mean FDE models for Internet congestion control under a many-flows regime , 2004, IEEE Transactions on Information Theory.

[14]  Cheng Song,et al.  High performance TCP in ANSNET , 1994, CCRV.

[15]  Donald F. Towsley,et al.  A study of networks simulation efficiency: fluid simulation vs. packet-level simulation , 2001, Proceedings IEEE INFOCOM 2001. Conference on Computer Communications. Twentieth Annual Joint Conference of the IEEE Computer and Communications Society (Cat. No.01CH37213).

[16]  Donald F. Towsley,et al.  Fluid models and solutions for large-scale IP networks , 2003, SIGMETRICS '03.

[17]  David M. Nicol,et al.  Discrete event fluid modeling of background TCP traffic , 2004, TOMC.

[18]  Michel Mandjes,et al.  Large Deviations for Small Buffers: An Insensitivity Result , 2001, Queueing Syst. Theory Appl..

[19]  Damon Wischik,et al.  SHRiNK: A method for scaleable performance prediction and efficient network simulation , 2003, IEEE INFOCOM 2003. Twenty-second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No.03CH37428).

[20]  João Pedro Hespanha,et al.  A hybrid systems modeling framework for fast and accurate simulation of data communication networks , 2003, SIGMETRICS '03.

[21]  G. Raina,et al.  Buffer sizes for large multiplexers: TCP queueing theory and instability analysis , 2005, Next Generation Internet Networks, 2005.

[22]  HighWire Press.  Philosophical Transactions of the Royal Society of London , 1781, The London Medical Journal.

[23]  François Baccelli,et al.  Flow level simulation of large IP networks , 2003, IEEE INFOCOM 2003. Twenty-second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No.03CH37428).

[24]  Steven H. Low,et al.  REM: active queue management , 2001, IEEE Netw..

[25]  Mostafa H. Ammar,et al.  Distributed network simulations using the dynamic simulation backplane , 2001, Proceedings 21st International Conference on Distributed Computing Systems.

[26]  Ugo Mocci,et al.  Broadband Network Teletraffic: Final Report of Action COST 242 , 1996 .

[27]  Donald F. Towsley,et al.  Unresponsive flows and AQM performance , 2003, IEEE INFOCOM 2003. Twenty-second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No.03CH37428).

[28]  Sanjay Shakkottai,et al.  Time-scale decomposition and equivalent rate-based marking , 2006, TNET.

[29]  Jennifer C. Hou,et al.  Network calculus based simulation for TCP congestion control: theorems, implementation and evaluation , 2004, IEEE INFOCOM 2004.

[30]  Guido Appenzeller,et al.  Sizing router buffers , 2004, SIGCOMM '04.

[31]  Donald F. Towsley,et al.  On integrating fluid models with packet simulation , 2004, IEEE INFOCOM 2004.

[32]  Jin Cao,et al.  A Poisson limit for buffer overflow probabilities , 2002, Proceedings.Twenty-First Annual Joint Conference of the IEEE Computer and Communications Societies.