Accelerating simulation of large-scale IP networks: A network invariant preserving approach

Hwangnam Kim, Hyuk Lim, Jennifer C. Hou

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

In this paper, we propose a simulation framework, TranSim, that reduces the rate at which packet-events are generated, in order to accelerate large-scale simulation of IP networks with TCP/UDP traffic. Conceptually, we transform an IP network into an alternate network that generates a smaller number of packet events, carry out simulation in the "transformed" network, and then extrapolate simulation results for the original network from those obtained in the "transformed" network. We formally prove that if the network invariant - the bandwidthdelay product -is preserved, the network dynamics, such as the queue dynamics and the packet dropping probability at each link, and TCP dynamics, such as the congestion window, RTTs, and rate dynamics, remain unchanged in the process of network transformation. We have implemented TranSim in ns-2, and conducted an empirical study to evaluate it against packet level simulation, with respect to the capability of capturing transient, packet level network dynamics, the reduction in the execution time and the memory usage, and the discrepancy in the system throughput. The simulation results indicate maximally two orders of magnitude improvement in the execution time and the performance improvement becomes more prominent as the network size increases (in terms of the number of nodes, the number of flows, the complexity of topology, and link capacity) or as the degree of downsizing increases. The memory usage incurred in TranSim is comparable to that in packet level simulation. The error discrepancy between TranSim and packet level simulation, on the other hand, is between 1-10 % in a wide variety of network topologies, inclusive of randomly generated topologies, and traffic loads with various AQM strategies.

Original languageEnglish
Title of host publicationProceedings - IEEE INFOCOM
DOIs
Publication statusPublished - 2006 Dec 1
Externally publishedYes
EventINFOCOM 2006: 25th IEEE International Conference on Computer Communications - Barcelona, Spain
Duration: 2006 Apr 232006 Apr 29

Other

OtherINFOCOM 2006: 25th IEEE International Conference on Computer Communications
CountrySpain
CityBarcelona
Period06/4/2306/4/29

Fingerprint

Topology
Data storage equipment
Throughput

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Hardware and Architecture

Cite this

Accelerating simulation of large-scale IP networks : A network invariant preserving approach. / Kim, Hwangnam; Lim, Hyuk; Hou, Jennifer C.

Proceedings - IEEE INFOCOM. 2006. 4146838.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kim, H, Lim, H & Hou, JC 2006, Accelerating simulation of large-scale IP networks: A network invariant preserving approach. in Proceedings - IEEE INFOCOM., 4146838, INFOCOM 2006: 25th IEEE International Conference on Computer Communications, Barcelona, Spain, 06/4/23. https://doi.org/10.1109/INFOCOM.2006.185
@inproceedings{1b46f5c4f3ec456cb3d6c07d686f84e9,
title = "Accelerating simulation of large-scale IP networks: A network invariant preserving approach",
abstract = "In this paper, we propose a simulation framework, TranSim, that reduces the rate at which packet-events are generated, in order to accelerate large-scale simulation of IP networks with TCP/UDP traffic. Conceptually, we transform an IP network into an alternate network that generates a smaller number of packet events, carry out simulation in the {"}transformed{"} network, and then extrapolate simulation results for the original network from those obtained in the {"}transformed{"} network. We formally prove that if the network invariant - the bandwidthdelay product -is preserved, the network dynamics, such as the queue dynamics and the packet dropping probability at each link, and TCP dynamics, such as the congestion window, RTTs, and rate dynamics, remain unchanged in the process of network transformation. We have implemented TranSim in ns-2, and conducted an empirical study to evaluate it against packet level simulation, with respect to the capability of capturing transient, packet level network dynamics, the reduction in the execution time and the memory usage, and the discrepancy in the system throughput. The simulation results indicate maximally two orders of magnitude improvement in the execution time and the performance improvement becomes more prominent as the network size increases (in terms of the number of nodes, the number of flows, the complexity of topology, and link capacity) or as the degree of downsizing increases. The memory usage incurred in TranSim is comparable to that in packet level simulation. The error discrepancy between TranSim and packet level simulation, on the other hand, is between 1-10 {\%} in a wide variety of network topologies, inclusive of randomly generated topologies, and traffic loads with various AQM strategies.",
author = "Hwangnam Kim and Hyuk Lim and Hou, {Jennifer C.}",
year = "2006",
month = "12",
day = "1",
doi = "10.1109/INFOCOM.2006.185",
language = "English",
isbn = "1424402212",
booktitle = "Proceedings - IEEE INFOCOM",

}

TY - GEN

T1 - Accelerating simulation of large-scale IP networks

T2 - A network invariant preserving approach

AU - Kim, Hwangnam

AU - Lim, Hyuk

AU - Hou, Jennifer C.

PY - 2006/12/1

Y1 - 2006/12/1

N2 - In this paper, we propose a simulation framework, TranSim, that reduces the rate at which packet-events are generated, in order to accelerate large-scale simulation of IP networks with TCP/UDP traffic. Conceptually, we transform an IP network into an alternate network that generates a smaller number of packet events, carry out simulation in the "transformed" network, and then extrapolate simulation results for the original network from those obtained in the "transformed" network. We formally prove that if the network invariant - the bandwidthdelay product -is preserved, the network dynamics, such as the queue dynamics and the packet dropping probability at each link, and TCP dynamics, such as the congestion window, RTTs, and rate dynamics, remain unchanged in the process of network transformation. We have implemented TranSim in ns-2, and conducted an empirical study to evaluate it against packet level simulation, with respect to the capability of capturing transient, packet level network dynamics, the reduction in the execution time and the memory usage, and the discrepancy in the system throughput. The simulation results indicate maximally two orders of magnitude improvement in the execution time and the performance improvement becomes more prominent as the network size increases (in terms of the number of nodes, the number of flows, the complexity of topology, and link capacity) or as the degree of downsizing increases. The memory usage incurred in TranSim is comparable to that in packet level simulation. The error discrepancy between TranSim and packet level simulation, on the other hand, is between 1-10 % in a wide variety of network topologies, inclusive of randomly generated topologies, and traffic loads with various AQM strategies.

AB - In this paper, we propose a simulation framework, TranSim, that reduces the rate at which packet-events are generated, in order to accelerate large-scale simulation of IP networks with TCP/UDP traffic. Conceptually, we transform an IP network into an alternate network that generates a smaller number of packet events, carry out simulation in the "transformed" network, and then extrapolate simulation results for the original network from those obtained in the "transformed" network. We formally prove that if the network invariant - the bandwidthdelay product -is preserved, the network dynamics, such as the queue dynamics and the packet dropping probability at each link, and TCP dynamics, such as the congestion window, RTTs, and rate dynamics, remain unchanged in the process of network transformation. We have implemented TranSim in ns-2, and conducted an empirical study to evaluate it against packet level simulation, with respect to the capability of capturing transient, packet level network dynamics, the reduction in the execution time and the memory usage, and the discrepancy in the system throughput. The simulation results indicate maximally two orders of magnitude improvement in the execution time and the performance improvement becomes more prominent as the network size increases (in terms of the number of nodes, the number of flows, the complexity of topology, and link capacity) or as the degree of downsizing increases. The memory usage incurred in TranSim is comparable to that in packet level simulation. The error discrepancy between TranSim and packet level simulation, on the other hand, is between 1-10 % in a wide variety of network topologies, inclusive of randomly generated topologies, and traffic loads with various AQM strategies.

UR - http://www.scopus.com/inward/record.url?scp=39049093620&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=39049093620&partnerID=8YFLogxK

U2 - 10.1109/INFOCOM.2006.185

DO - 10.1109/INFOCOM.2006.185

M3 - Conference contribution

AN - SCOPUS:39049093620

SN - 1424402212

SN - 9781424402212

BT - Proceedings - IEEE INFOCOM

ER -