DocumentCode :
1827958
Title :
A very accurate approximation for cell loss ratio in ATM networks
Author :
Haghighat, A.T. ; Faez, K.
Author_Institution :
Dept. of Comput. Eng., Amirkabir Univ. of Technol., Tehran, Iran
fYear :
2001
fDate :
10-12 Oct. 2001
Firstpage :
282
Lastpage :
286
Abstract :
We like to find the cell loss ratio in ATM networks when the statistical multiplexing is an important factor. In this paper, first we have proposed the combination of three analytical expressions, which approximate the cell loss probability, based on the fluid-flow approximation model and two stationary approximation models. Second, we have provided a very accurate numerical model for the finite buffer, which lies at the input of each VP. The sources are statistically independent and each traffic source has a two-state Markov model. This simulation is done at the cell level and its results are very accurate. We have compared the results of the numerical simulation with the results of the analytical approximation models. Also we have used linear estimation to find an accurate expression for cell loss approximation in ATM networks.
Keywords :
Markov processes; approximation theory; asynchronous transfer mode; buffer storage; probability; quality of service; telecommunication network routing; telecommunication traffic; ATM networks; ATM routing algorithm; QoS; accurate approximation; accurate numerical model; analytical approximation models; cell level simulation; cell loss approximation; cell loss probability; cell loss ratio; finite buffer; fluid-flow approximation model; linear estimation; numerical simulation; quality of service; stationary approximation models; statistical multiplexing; traffic source; two-state Markov model; virtual path; Analytical models; Asynchronous transfer mode; Bit rate; Computer networks; Intelligent networks; Numerical models; Probability; Propagation losses; Quality of service; Traffic control;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Networks, 2001. Proceedings. Ninth IEEE International Conference on
ISSN :
1531-2216
Print_ISBN :
0-7695-1187-4
Type :
conf
DOI :
10.1109/ICON.2001.962354
Filename :
962354
Link To Document :
بازگشت