Traffic load balancing in low Earth orbit satellite networks
Authors:
- Young-Ho BaeYBYoung-Ho Bae
- Youngjae KimYKYoungjae Kim
- Chul Hye ParkCPChul Hye Park
- 27 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 10952055,p. 191-195
Abstract
One of the unique traffic features of low Earth orbit (LEO) satellite networks is the time-variant and nonuniform load distribution. This feature results in a locally biased congestion problem for the LEO satellite systems. We propose a traffic load balancing scheme to resolve the congestion problem in such traffic scenarios. The proposing scheme makes use of near-neighbor residual bandwidth information to apportion excess bandwidth from congested satellites to their underloaded neighbors in the network. Each traffic load balancing process is performed on the domain basis. When the residual bandwidth of the path by way of the target satellite is less than a prespecified amount, calls routed over the path can be migrated to the alternative paths of the same balancing domain. By decreasing the size of the network graph being considered in load balancing process, balancing domains reduce the complexity of the process.Keywords
This publication has 8 references indexed in Scilit:
- Managing connection handover in satellite networks2002
- A connection handover protocol for LEO satellite ATM networks1997
- ATM-based routing in LEO/MEO satellite networks with intersatellite linksIEEE Journal on Selected Areas in Communications, 1997
- A dynamic routing concept for ATM-based satellite personal communication networksIEEE Journal on Selected Areas in Communications, 1997
- Traffic characteristics of LEOS-based global personal communications networksIEEE Communications Magazine, 1997
- Performance of an integrated voice/data system in nonuniform traffic low Earth-orbit satellite communication systemsIEEE Journal on Selected Areas in Communications, 1995
- Mobile satellite communicationsProceedings of the IEEE, 1994
- Strategies for dynamic load balancing on highly parallel computersIEEE Transactions on Parallel and Distributed Systems, 1993