Leveraging Degree Centrality for Fast Bandwidth Measurement in OpenFlow-based SDN

Network optimization and policy deployment need the underlying state information as a reference basis. However, the timeliness of information collection will seriously affect the effectiveness of optimization results. As a key index, link bandwidth has a guiding significance for network management such as routing, queue scheduling and congestion control. In this paper, we use the theory of degree centrality-based critical node ranking in complex networks to implement fast link bandwidth measurements in software-defined networks. Simulation results on several real topologies of different sizes verify the advantages of our scheme over existing algorithms in terms of measurement time overhead and robustness to different measurement methods.