You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The network is modelled as an undirected graph G(V,E) by using the NetworkX tool, https://networkx.github.io/. To represent the data plane topology, Abilene network topology has been adopted from http://sndlib.zib.de/home.action. 6 clients are added to generate traffic and represent different class of services.
Fig.3: Abilene Topology
Experimental Results:
Fig.4: Evaluation of two different CoS flows: video and ICMP.
Information is captured from the {s 2 , s 6 } link of the Abilene topology. The channel capacity is 8Mbps. The video and ICMP flows traverse a common path (s 1 , s 2 , s 6 , s 8 , s 7 , s 9 ).
Fig.5: Network overhead measurement: the total number of
invoked OpenFlow messages are: 2578, 2735 and 2764 messages for
GMSM, active-flow and periodic monitoring respectively.
Fig.6: The count of OpenFlow stats request/reply packets per 1ms
versus time are illustrated when the monitoring function is set at 10s
intervals for GMSM, Active Flow and Periodic Monitoring. The □
represents the stats request and ♢ represents stats reply.
Fig.7: Number of switches used in monitoring: GMSM queries
the smallest number of switches vs periodic and active monitoring.
- `If you use this framework or any of its code in your work then, please cite our IEEE Access publication:+ A.Malik,R.de Fréin,"Graph Modeling for OpenFlow Switch Monitoring",IEEE Access,vol.11,p.84543-84553,2023" `