Subscribe Now Subscribe Today
 
FOLLOW US:     Facebook     Twitter
   
Journal of Software Engineering
  Year: 2014 | Volume: 8 | Issue: 4 | Page No.: 304-313
DOI: 10.3923/jse.2014.304.313
An OpenFlow Switch Model using Kepler GPU
Li Xiuhai

Abstract:
OpenFlow makes up for traditional network, but network session identification is inefficient and packet forwarding path selection is poor. Focusing on forwarding path and matching; we propose OGMD model which combin GPU with biological sequence algorithms and machine learning methods to accelerate matching speed and to improve network environment, we also proposed matching algorithm and path selection algorithm. Experiments show matching algorithm gives a speedup of 325 and path selection algorithm makes link loss rate less than 5%, with an average decline 69.35% and network delay less than 20 msec, average fell 63.28%.
 [Fulltext PDF]   [Fulltext HTML]   [XML: Abstract + References]   [References]   [Report Citation]
How to cite this article:

Li Xiuhai , 2014. An OpenFlow Switch Model using Kepler GPU. Journal of Software Engineering, 8: 304-313.

DOI: 10.3923/jse.2014.304.313

URL: http://scialert.net/abstract/?doi=jse.2014.304.313

 
COMMENT ON THIS PAPER
.