The Community for Technology Leaders
2016 International Conference on Computing, Networking and Communications (ICNC) (2016)
Kauai, HI, USA
Feb. 15, 2016 to Feb. 18, 2016
ISBN: 978-1-4673-8578-7
pp: 1-6
Lilin Zhang , The Edward S. Rogers Sr. Department of Electrical & Computer Engineering, University of Toronto, Ontario, Canada
Ali Tizghadam , The Edward S. Rogers Sr. Department of Electrical & Computer Engineering, University of Toronto, Ontario, Canada
Hadi Bannazadeh , The Edward S. Rogers Sr. Department of Electrical & Computer Engineering, University of Toronto, Ontario, Canada
Alberto Leon-Garcia , The Edward S. Rogers Sr. Department of Electrical & Computer Engineering, University of Toronto, Ontario, Canada
ABSTRACT
The quality of multimedia communications heavily relies on the end-to-end network condition. Media sessions that are routed through highly-utilized links are in more jeopardy of longer delays, more packet loss. Choosing optimal routes to embed the data flows can be performed in a centralized manner in SDN-enabled networks. However, the calculation of such optimal flow embedding is NP-hard. In this paper, we propose a Two-phase Flow Embedding heuristic to tackle the problem. Phase I is a global planning module, periodically invoked. Given the total traffic of a time period, it produces the optimal end-to-end tunnels between each pair of communication endpoints. The objective is to balance the traffic distribution, measured by the value of network criticality. Phase II is a traffic engineering module, always active. It receives the individual data flows, and selects one of the pre-configured tunnels between the requested endpoints, to embed the flow. We compare the Two-phase flow embedding heuristic with Dijkstra algorithm. We show that the proposed heuristic outperforms the Dijkstra's in three different metrics: session accept rate, maximum/average substrate link utilization, and network criticality.
INDEX TERMS
CITATION

L. Zhang, A. Tizghadam, H. Bannazadeh and A. Leon-Garcia, "Flow embedding in the data plane of multimedia IP communications," 2016 International Conference on Computing, Networking and Communications (ICNC), Kauai, HI, USA, 2016, pp. 1-6.
doi:10.1109/ICCNC.2016.7440664
345 ms
(Ver 3.3 (11022016))