QoS provisioning for TT-TR based OBS ring networks

Limei Peng, Young Chon Kim, Wan Tang, Chan Hyun Youn

Research output: Contribution to journalArticlepeer-review

Abstract

Optical burst switching (OBS) is one of the most promising switching technologies. This paper studies the efficient OBS ring network using the TT-TR based node architecture. The TT-TR based OBS ring network can use the network resources very efficiently, but the network performance is restricted due to the limited network resources. Moreover, different network demands have different network resource requirements. It is very important to satisfy their different requirements while using the limited network resources efficiently. In this paper, Quality of Service (QoS) provisioning mechanisms are studied to solve the above mentioned problems. Two kinds of network services, namely the real-time traffic and the best-effort traffic, are considered in this paper. The real-time service is delay-sensitive that should be transmitted in a real-time manner and cannot tolerate high transmission delay; while the best-effort traffic has no strict limitation in transmission delay or loss, but the best effort are made to improve their transmission reliability. According to their different servicing requirements, different QoS provisioning mechanisms are proposed to serve them in the OBS unidirectional and bidirectional ring networks, respectively. The proposed mechanisms are evaluated and analyzed through simulation. The simulation results show that even though the proposed mechanisms cannot get absolute zero-delay or zero-loss for real-time traffic or best-effort traffic, their different requirements for delay or loss are satisfied to some significant degree.

Original languageEnglish
Pages (from-to)117-125
Number of pages9
JournalJournal of Optical Communications
Volume32
Issue number2
DOIs
StatePublished - 2011

Keywords

  • Optical burst switching
  • Quality of service
  • TT-TR

Fingerprint

Dive into the research topics of 'QoS provisioning for TT-TR based OBS ring networks'. Together they form a unique fingerprint.

Cite this