We evaluate and compare the performance of timer-based and threshold-based assembly algorithms in Optical Burst Switching networks. Results including burst blocking probability, mean packet delay and link utilization at the ingress node are presented from both simulations and two theoretical models. The results are obtained for the full range of input traffic load so they can provide guidelines for design and dimensioning links to meet desired Quality of Service levels.
This letter proposes a new dynamic burst length adjustment mechanism to achieve optimal performance of fiber delay line (FDL) buffers in optical burst switching networks. Our results show that burst blocking probability at an optical node with a fixed length FDL buffer can be kept at minimum under changing traffic conditions using the proposed mechanism.
Abstract:The sending rate of commonly used TCP protocols is tightly coupled to packet loss within the network: a high rate of packet loss will cause a sender to slow down, thereby reducing the network load and decreasing subsequent packet loss rates. In this paper, we combine a widely verified source rate TCP model with an Optical Burst Switching (OBS) loss model, to find fixed-point input loads and loss rates for an OBS link carrying TCP traffic. In doing so, we show that if OBS networks are to be efficiently used to carry TCP traffic, many wavelengths with full wavelength conversion are required.
We propose a cost effective protection architecture and its algorithms in the OLT (Optical Line Termination) and the ONUs (Optical Network Units) for Ethernet Passive Optical Network (EPON). The proposed protection architecture can accommodate various types of protection demands of subscribers. We show that our protection mechanism can recover within 50111s. Finally we prove the proposed architecture as a cost effective solution compared with other similar protection architectures Kepvordr-EPON; protection architecture; access nehvorh
Legacy data centers are transforming toward cloud data centers according to the advance of mobile and Internet of Things technology, processing of big data, and development of cloud computing technology. The goal of cloud data centers is to efficiently manage energy and facility, and to rapidly provide service demands to users by operating virtualized ICT(Information and Communication Technology) resources. Accordingly, it requires to configure and operate networks for efficiently providing virtualized ICT resources. This paper analyzes networking technologies suitable for cloud data centers and presents ways to efficiently operate the data center.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.