Details
| Original language | English |
|---|---|
| Title of host publication | 2018 IFIP Networking Conference IFIP Networking and Workshops |
| Subtitle of host publication | IFIP Networking 2018 - Proceedings |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 460-468 |
| Number of pages | 9 |
| ISBN (electronic) | 9783903176089 |
| Publication status | Published - 2 Jul 2018 |
| Event | 17th IFIP Networking Conference IFIP Networking and Workshops, IFIP Networking 2018 - Zurich, Switzerland Duration: 14 May 2018 → 16 May 2018 |
Abstract
The dispersion that arises when packets traverse a network carries information that can reveal relevant network characteristics. Using a fluid-flow model of a bottleneck link with first-in first-out multiplexing, accepted probing tools measure the packet dispersion to estimate the available bandwidth, i.e., the residual capacity that is left over by other traffic. Difficulties arise, however, if the dispersion is distorted compared to the model, e.g., by non-fluid traffic, multiple bottlenecks, clustering of packets due to interrupt coalescing, and inaccurate time-stamping in general. It is recognized that modeling these effects is cumbersome if not intractable. This motivates us to explore the use of machine learning in bandwidth estimation. We train a neural network using vectors of the packet dispersion that is characteristic of the available bandwidth. Our testing results reveal that even a shallow neural network identifies the available bandwidth with high precision. We also apply the neural network under a variety of notoriously difficult conditions that have not been included in the training, such as heavy traffic burstiness, and multiple bottleneck links. Compared to two state-of-the-art model-based techniques, the neural network approach shows improved performance. Further, the neural network can effectively control the estimation procedure in an iterative implementation.
ASJC Scopus subject areas
- Computer Science(all)
- Computer Networks and Communications
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
2018 IFIP Networking Conference IFIP Networking and Workshops: IFIP Networking 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2018. p. 460-468 8697023.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Neural Networks for Measurement-based Bandwidth Estimation
AU - Khangura, Sukhpreet Kaur
AU - Fidler, Markus
AU - Rosenhahn, Bodo
PY - 2018/7/2
Y1 - 2018/7/2
N2 - The dispersion that arises when packets traverse a network carries information that can reveal relevant network characteristics. Using a fluid-flow model of a bottleneck link with first-in first-out multiplexing, accepted probing tools measure the packet dispersion to estimate the available bandwidth, i.e., the residual capacity that is left over by other traffic. Difficulties arise, however, if the dispersion is distorted compared to the model, e.g., by non-fluid traffic, multiple bottlenecks, clustering of packets due to interrupt coalescing, and inaccurate time-stamping in general. It is recognized that modeling these effects is cumbersome if not intractable. This motivates us to explore the use of machine learning in bandwidth estimation. We train a neural network using vectors of the packet dispersion that is characteristic of the available bandwidth. Our testing results reveal that even a shallow neural network identifies the available bandwidth with high precision. We also apply the neural network under a variety of notoriously difficult conditions that have not been included in the training, such as heavy traffic burstiness, and multiple bottleneck links. Compared to two state-of-the-art model-based techniques, the neural network approach shows improved performance. Further, the neural network can effectively control the estimation procedure in an iterative implementation.
AB - The dispersion that arises when packets traverse a network carries information that can reveal relevant network characteristics. Using a fluid-flow model of a bottleneck link with first-in first-out multiplexing, accepted probing tools measure the packet dispersion to estimate the available bandwidth, i.e., the residual capacity that is left over by other traffic. Difficulties arise, however, if the dispersion is distorted compared to the model, e.g., by non-fluid traffic, multiple bottlenecks, clustering of packets due to interrupt coalescing, and inaccurate time-stamping in general. It is recognized that modeling these effects is cumbersome if not intractable. This motivates us to explore the use of machine learning in bandwidth estimation. We train a neural network using vectors of the packet dispersion that is characteristic of the available bandwidth. Our testing results reveal that even a shallow neural network identifies the available bandwidth with high precision. We also apply the neural network under a variety of notoriously difficult conditions that have not been included in the training, such as heavy traffic burstiness, and multiple bottleneck links. Compared to two state-of-the-art model-based techniques, the neural network approach shows improved performance. Further, the neural network can effectively control the estimation procedure in an iterative implementation.
UR - http://www.scopus.com/inward/record.url?scp=85065503619&partnerID=8YFLogxK
U2 - 10.23919/ifipnetworking.2018.8697023
DO - 10.23919/ifipnetworking.2018.8697023
M3 - Conference contribution
AN - SCOPUS:85065503619
SP - 460
EP - 468
BT - 2018 IFIP Networking Conference IFIP Networking and Workshops
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 17th IFIP Networking Conference IFIP Networking and Workshops, IFIP Networking 2018
Y2 - 14 May 2018 through 16 May 2018
ER -