Browse Prior Art Database

Predicting Link Load Using Extrapolation and Gradient Measurement

IP.com Disclosure Number: IPCOM000028690D
Published in the IP.com Journal: Volume 4 Issue 6 (2004-06-25)
Included in the Prior Art Database: 2004-Jun-25
Document File: 1 page(s) / 27K

Publishing Venue

Siemens

Related People

Juergen Carstens: CONTACT

Abstract

Source nodes in communication networks need information about the available bandwidth of its links to other nodes to choose the best route for the transmitted signal and to avoid over-utilization of specific links. At present the best information a source node has about the condition of the network comes from triggered or normal Link State Advertisement (LSA) that contain information about available bandwidth of a link. Unfortunately this information just reflects the current situation and may be outdated after some time. Therefore, it is hereby proposed to use an extrapolation algorithm based on the recent updates to predict the available bandwidth until a new update is received (figure 1). The algorithm can be of first (linear) or higher degree. This reduces the risk of a route setup failure due to outdated information and means more accurately calculated routes.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 82% of the total text.

Page 1 of 1

S

Predicting Link Load Using Extrapolation and Gradient Measurement

Idea: Richard Edwin, GB-Romsey; Andrew Reeve, GB-Romsey; Changpeng Fan, DE-Berlin

Source nodes in communication networks need information about the available bandwidth of its links to other nodes to choose the best route for the transmitted signal and to avoid over-utilization of specific links. At present the best information a source node has about the condition of the network comes from triggered or normal Link State Advertisement (LSA) that contain information about available bandwidth of a link. Unfortunately this information just reflects the current situation and may be outdated after some time.

Therefore, it is hereby proposed to use an extrapolation algorithm based on the recent updates to predict the available bandwidth until a new update is received (figure 1). The algorithm can be of first (linear) or higher degree. This reduces the risk of a route setup failure due to outdated information and means more accurately calculated routes.

Additionally, it is proposed to record the available bandwidth of a link and constantly calculate the rate of change (gradient) so a rapid reduction of the available bandwidth can be detected early (figure 2). When such a fast reduction is measured, the weighting for that link is adjusted so that the routing algorithms avoid or back off from their usage of that link. This will bring the curve back to a more reasonable gradient and so avoid over-utilizatio...