An Adaptive Traffic Distribution Scheme for CMT based on Lotka-Volterra Model in Multihomed Networks
CMT (concurrent multipath trans- fer) can increase throughput and transmission efficiency in multihomed networks. However, it is still an important challenge about distributing traffic adaptively into multiple access networks. Based on LV (Lotka-Volterra) model, we propose an adaptive traffic distri...
Gespeichert in:
Veröffentlicht in: | China communications 2017-02, Vol.14 (2), p.79-89 |
---|---|
Hauptverfasser: | , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | CMT (concurrent multipath trans- fer) can increase throughput and transmission efficiency in multihomed networks. However, it is still an important challenge about distributing traffic adaptively into multiple access networks. Based on LV (Lotka-Volterra) model, we propose an adaptive traffic distribution scheme. In the scheme, two com- petition modes are concluded, multiple S-D (source-destination) streams competition for bandwidth of one path and multipath competi- tion for traffic between each multihomed S-D host pair. Actually, each access network can establish a path for S-D pairs. So, in the first mode, each path is analogous to a predator, and overall traffic in a multihomed host is analogous to prey. Then, each path has to com- pete for the traffic by path information, e.g., bandwidth and congestion level. In the other one, if several S-D pairs pass through a shared path simultaneously, they will compete for bandwidth of the path. Here, each S-D pair is analogous to a predator, and the bandwidth of the common path is analogous to the prey. At last, compared with other three schemes, uniform traffic distribution, greedy path selection, random path selection in OPNET simulator, the proposed scheme can perform better on reducing file transmission time and increasing network throughput in FTP service. |
---|---|
ISSN: | 1673-5447 |
DOI: | 10.1109/CC.2017.7868177 |