|Title:||CBR QoS measurement in ATM networks using a programmed ATM generator|
|Subject:||Asynchronous transfer mode|
Hong Kong Polytechnic University -- Dissertations
Department of Electronic and Information Engineering
|Pages:||v, 88 leaves : ill. ; 30 cm|
|Abstract:||In this project, we measure the Cell Transfer Delay (CTD1) and Cell Delay Variation (CDV2) of Constant Bit Rate (CBR) traffic at 25.6 Mbps. This project is a continuation of the MSc dissertation work of Mr. Sung Sai Ping, a former MSc student of Dr. Alex Wai. In his work, Mr. Sung used the build-in test cell streams in the Asynchronous Transfer Mode (ATM3) tester at 155 Mbps to measure the effect of buffer contention between the foreground and background traffic on the CTD and CDV of the foreground cell stream. He found that CTD and CDV are almost constant at about half a cell time until the background load reaches about 90%. At that point, the CTD and CDV rise sharply to several cell times. Also the CTD and CDV increase as the number of hops increases. In this project, an ATM CBR generator program was written in Visual C++ using Winsock 2 API. An ATM Testbed was set up at 25.6Mbps instead of 155 Mbps. The software CBR generator is used to generate the background traffic while an Adtech ATM tester is used to generate the reference foreground traffic. In each of the output ports, both the foreground and background traffic streams compete for the buffer as they exit from the ATM switch. The CDV and (CTD of an ATM network transporting CBR traffic at 25.6 Mbps was then measured. We evaluated the performance of CDV and CTD under different network parameters such as the number of switches transverse and the background loading. We found that the results are qualitatively different from that obtained using ATM-155 OC-3 set up  and . The sensitivity of CTD and CDV in ATM-25 is much lower than that of ATM-155. In addition, we studied the distribution of the measured CTD and find that it is Gaussian (Normal) distributed for larger number of hops and high traffic loading under non-synchronized traffic streams. We then replace the ATM program generator by the Adtech ATM CBR generator and repeat the same measurement. The results show that the behaviour of the mean CTD, the standard deviations of the CTD and CDV are different from that obtained in  and . The CTD density function from the Adtech Test Set generator is not Gaussian distributed. This indicates that using test cell steams from the Adtech analyzer in the above CDV and CDT measurement is not suitable because the background traffic streams at each buffer are synchronized.|
|Rights:||All rights reserved|
Files in This Item:
|b15383301.pdf||For All Users (off-campus access for PolyU Staff & Students only)||2.75 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: