.png){kind=icon}
.png){kind=icon}
Return
Distributed Strategy for Collaborative Traffic Measurement in a Multi-Controller SDN
.png){kind=icon}
PRE
AI
DOI：10.1109/TNSE.2023.3271123.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Traffic measurement provides fundamental flow statistics for a wide range of network applications. In traditional networks, since switches work independently, a flow may be measured by multiple switches along its route path, which is a waste of network resource. Software defined networking (SDN) is a new paradigm that decouples control plane and data plane. The control plane is implemented in a centralized controller to manage all switches with a global view. Many works focus on designing efficient centralized strategy to coordinate the switches for traffic measurement, called collaborative traffic measurement. However, with the continuous growth of the network topology and traffic volume, the controller may be the bottleneck. To improve the capability and enhance the robustness of control plane, multiple controllers is employed to manage a subset of switches, respectively. Unfortunately, control plane with multiple distributed controllers cannot execute the centralized measurement strategy. In this paper, we propose a novel distributed iterative strategy for collaborative traffic measurement in a multi-controller SDN. We also theoretically prove that our proposed algorithm will converge to the global optimal by iterations. The extensive simulations demonstrate that the proposed strategy achieves a near-optimal performance (gap about 10%) in terms of measurement load at switches and drastically reduces the communication load by 20% at least among controllers compared with the state-of-the-art algorithms.
Keywords:
Control systems
Collaboration
Coordinate measuring machines
Termination of employment
Software
Size measurement
Redundancy
Collaborative traffic measurement
multiple controllers
software defined networks
Journal
I
IF:
7.9
Papers:
2.5K
Citations:
10.0K