Loading...

Traffic Engineering In Cloud Infrastructure By Software Defined Networks

Karami, Saeedeh | 2015

608 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: English
  3. Document No: 48460 (52)
  4. University: Sharif University of Technology, International Campus, Kish Island
  5. Department: Science and Engineering
  6. Advisor(s): Akbari, Behzad
  7. Abstract:
  8. In the recent years, cloud data center issues have been the center of attention. Also for them topologies with multi paths between end hosts are offered. For use of multi path traffic engineering is essential. Researches show congestion occurred in the cloud data center networks, because the mean of link utilization is down. This means, some of links are congested while there are some other with much available bandwidth. Currently, the most common method is Equal Cost Multipath (ECMP) algorithm for routing flows in the data center network. ECMP decides on routing without any attention to network congestion condition. There, may be two big flows placed in the one path and it causes creation of congestion and avoidable bottleneck. Other central methods such as Hedera , Mahout and MicroTE do routing according to flows size and/ or traffic forecast that these methods are inappropriate for dynamic and unpredictable traffic pattern. In this thesis a method for cloud data center networks traffic engineering by software defined networks has been suggested. This method is not according to traffic pattern forecast or routing with flow size. This method does traffic engineering with finding the shortest path according to the load volume on the paths. This method is implemented using with OpenFlow protocol. Evaluation of proposed method shows that this method has improvement 11.02 percent for the maximum link utilization(MLU) than ECMP algorithms
  9. Keywords:
  10. Traffic Engineering ; Cloud Computing ; Software Defined Networking (SDN) ; Multipath Routing ; OpenFlow Protocol

 Digital Object List

 Bookmark

No TOC