Analytical models for an Ethernet-like local area network link

Authors:
Madhav Marathe;Sujit Kumar
Affiliations:
Digital Equipment Corporation, 146 Main Street, Maynard, Massachusetts;Digital Equipment Corporation, 146 Main Street, Maynard, Massachusetts
Venue:
SIGMETRICS '81 Proceedings of the 1981 ACM SIGMETRICS conference on Measurement and modeling of computer systems
Year:
1981

Citing 3
Cited 3

Measured performance of an Ethernet local network

Communications of the ACM
Ethernet: distributed packet switching for local computer networks

Communications of the ACM
The behavior of Ethernet-like computer communications networks

SOSP '79 Proceedings of the seventh ACM symposium on Operating systems principles

Response Time Preservation: A general technique for developing approximate algorithms for queueing networks

SIGMETRICS '84 Proceedings of the 1984 ACM SIGMETRICS conference on Measurement and modeling of computer systems
A virtual server queueing network method for component based performance modelling of metacomputing

Future Generation Computer Systems - Special issue: Semantic grid and knowledge grid: the next-generation web
Measurement and Analysis of HYPERchannel Networks

IEEE Transactions on Computers

Quantified Score

Hi-index	0.00

Visualization

Abstract

Ethernet -like local area network links have been studied by a number of researchers. Most of these studies have involved extensive simulation models operating at the level of individual packets. However, as we begin building models of systems built around such links, detailed simulation models are neither necessary, nor cost-effective. Instead, a simple analytical model of the medium should be adequate as a component of the higher level system models. This paper discusses a number of analytical models and identifies a last-in-first-out M/G/1 model with slightly increased service time as one which adequately captures both the mean and the coefficient of variation of the response time. Given any offered load, this model can be used to predict the mean waiting time and its coefficient of variation. These two can be used to construct a suitable 2 stage hyperexponential distribution. Random numbers can then be drawn from this distribution for use as waiting times of individual packets.