Search Results

1. H. Zhu, Comparison Between Pre-Computation and On-Demand Computation QoS Routing with Different Link State Update Algorithms, Master's Thesis, Networking Laboratory, Helsinki University of Technology, 2004 (bib)

   Abstract: As an extension to the current Best Effort service in the Internet, QoS routing may play an important role in providing quality guaranteed service. Extra cost is introduced by QoS routing. The cost mainly includes path computation cost and the cost for maintaining a QoS routing table.\ This thesis begins with a brief introduction to the QoS routing, and then the issue of inaccurate routing information and its impact on network performance are discussed. After that, the pre-computation routing algorithm and the on-demand computation routing algorithm are presented, followed by the comparison between these two algorithms.\ In order to run the simulations for this thesis, extensions to QRS simulator were made by adding new components and a Traffic Generator. Two new routing algorithms were added to simulate the pre-computation and the on-demand computation algorithms, respectively. By running simulations for these two algorithms with different link state update algorithms, the network performance is analyzed. Four different periods in PB and four thresholds in TB have been chosen for the simulations. Based on the simulation results, a conclusion is drawn: the frequency of link state update can affect the network performance and it costs significantly. Smaller link state update period or threshold can improve the network performance by reducing the inaccuracy in routing information, but the cost will also increase with more frequent link state updates. Thus, the update triggering policy should be chosen carefully.