Burst assembly at edge nodes is an important issue for the Optical Burst Switching (OBS) networks because it has a great impact on the traffic characteristics. We analyze the assembled traffic of the Science Information Network (SINET) by using the Fractional Brownian Motion (FBM) model. The analytical and simulation results show that existing assembly schemes cannot avoid increasing the burstiness, which will deteriorate the network performance. Here, burstiness is defined as the variance of the bitrate in small timescales. Therefore, we address the issue of how to reduce the burstiness of the assembled network traffic. Firstly, a sliding window-based assembly algorithm is introduced to reduce the burstiness of assembled traffic by transmitting bursts at an average rate in a small timescale. Next, an advanced timer-based assembly algorithm is introduced, by which the traffic rate is smoothed out by restricting the burst length to a threshold. The simulation results show that both the sliding window-based and advanced timer-based assembly algorithms perform better than existing assembly algorithms do in terms of the burst loss ratio. The simulation also indicates that the advanced timer-based assembly algorithm performs better in terms of the edge buffering delay than the sliding window-based assembly algorithm does.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Ping DU, Shunji ABE, "Traffic Analysis and Traffic-Smoothing Burst Assembly Methods for the Optical Burst Switching Network" in IEICE TRANSACTIONS on Communications,
vol. E90-B, no. 7, pp. 1620-1630, July 2007, doi: 10.1093/ietcom/e90-b.7.1620.
Abstract: Burst assembly at edge nodes is an important issue for the Optical Burst Switching (OBS) networks because it has a great impact on the traffic characteristics. We analyze the assembled traffic of the Science Information Network (SINET) by using the Fractional Brownian Motion (FBM) model. The analytical and simulation results show that existing assembly schemes cannot avoid increasing the burstiness, which will deteriorate the network performance. Here, burstiness is defined as the variance of the bitrate in small timescales. Therefore, we address the issue of how to reduce the burstiness of the assembled network traffic. Firstly, a sliding window-based assembly algorithm is introduced to reduce the burstiness of assembled traffic by transmitting bursts at an average rate in a small timescale. Next, an advanced timer-based assembly algorithm is introduced, by which the traffic rate is smoothed out by restricting the burst length to a threshold. The simulation results show that both the sliding window-based and advanced timer-based assembly algorithms perform better than existing assembly algorithms do in terms of the burst loss ratio. The simulation also indicates that the advanced timer-based assembly algorithm performs better in terms of the edge buffering delay than the sliding window-based assembly algorithm does.
URL: https://global.ieice.org/en_transactions/communications/10.1093/ietcom/e90-b.7.1620/_p
Copy
@ARTICLE{e90-b_7_1620,
author={Ping DU, Shunji ABE, },
journal={IEICE TRANSACTIONS on Communications},
title={Traffic Analysis and Traffic-Smoothing Burst Assembly Methods for the Optical Burst Switching Network},
year={2007},
volume={E90-B},
number={7},
pages={1620-1630},
abstract={Burst assembly at edge nodes is an important issue for the Optical Burst Switching (OBS) networks because it has a great impact on the traffic characteristics. We analyze the assembled traffic of the Science Information Network (SINET) by using the Fractional Brownian Motion (FBM) model. The analytical and simulation results show that existing assembly schemes cannot avoid increasing the burstiness, which will deteriorate the network performance. Here, burstiness is defined as the variance of the bitrate in small timescales. Therefore, we address the issue of how to reduce the burstiness of the assembled network traffic. Firstly, a sliding window-based assembly algorithm is introduced to reduce the burstiness of assembled traffic by transmitting bursts at an average rate in a small timescale. Next, an advanced timer-based assembly algorithm is introduced, by which the traffic rate is smoothed out by restricting the burst length to a threshold. The simulation results show that both the sliding window-based and advanced timer-based assembly algorithms perform better than existing assembly algorithms do in terms of the burst loss ratio. The simulation also indicates that the advanced timer-based assembly algorithm performs better in terms of the edge buffering delay than the sliding window-based assembly algorithm does.},
keywords={},
doi={10.1093/ietcom/e90-b.7.1620},
ISSN={1745-1345},
month={July},}
Copy
TY - JOUR
TI - Traffic Analysis and Traffic-Smoothing Burst Assembly Methods for the Optical Burst Switching Network
T2 - IEICE TRANSACTIONS on Communications
SP - 1620
EP - 1630
AU - Ping DU
AU - Shunji ABE
PY - 2007
DO - 10.1093/ietcom/e90-b.7.1620
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E90-B
IS - 7
JA - IEICE TRANSACTIONS on Communications
Y1 - July 2007
AB - Burst assembly at edge nodes is an important issue for the Optical Burst Switching (OBS) networks because it has a great impact on the traffic characteristics. We analyze the assembled traffic of the Science Information Network (SINET) by using the Fractional Brownian Motion (FBM) model. The analytical and simulation results show that existing assembly schemes cannot avoid increasing the burstiness, which will deteriorate the network performance. Here, burstiness is defined as the variance of the bitrate in small timescales. Therefore, we address the issue of how to reduce the burstiness of the assembled network traffic. Firstly, a sliding window-based assembly algorithm is introduced to reduce the burstiness of assembled traffic by transmitting bursts at an average rate in a small timescale. Next, an advanced timer-based assembly algorithm is introduced, by which the traffic rate is smoothed out by restricting the burst length to a threshold. The simulation results show that both the sliding window-based and advanced timer-based assembly algorithms perform better than existing assembly algorithms do in terms of the burst loss ratio. The simulation also indicates that the advanced timer-based assembly algorithm performs better in terms of the edge buffering delay than the sliding window-based assembly algorithm does.
ER -