Title: An improved trusted method for global congestion price computing based on software defined networking in data-centred network
Authors: Shan Chun; Chen Xiao-Long
Addresses: School of Electronics and Information, Guangdong Polytechnic Normal University, Guangzhou Guangdong, 510665, China ' School of Information Engineer, Jinhua Polytechnic, Jinhua Zhejiang, 321000, China
Abstract: The past computing method of the link status price is difficult to meet the needs of dynamic network. This paper is presented to take full advantage of the feature that the global link status information can be grasped by the central controller of software defined networking (SDN) architecture and proposed a two-tier trusted method for global congestion price computing (GCPC). The ratio argument vector B is given by the upper-tier method with the machine learning algorithm and fuzzy C-means clustering algorithm according to the global network operating status information. The link status price is computed according to ratio argument vector B. The simulation results show that the link status price calculated by GCPC method is trusted and this improved method can efficiently promote the throughput of bisection bandwidth.
Keywords: data-centred; software defined networking; the trusted price; global link status; congestion control.
International Journal of Intelligent Information and Database Systems, 2018 Vol.11 No.2/3, pp.153 - 168
Received: 08 Sep 2017
Accepted: 04 Nov 2017
Published online: 22 May 2018 *