蔡志宏臺灣大學：電信工程學研究所駱怡榮Lo, Yi-jungYi-jungLo2007-11-272018-07-052007-11-272018-07-052004http://ntur.lib.ntu.edu.tw//handle/246246/58624As Internet grows rapidly and bandwidth increases, ISP or CDN operators are able to provide high-data-rate applications for more and more customers. But owing to the drawbacks of the original Internet best effort design, such network will often be in chaos if we do not employ any QoS mechanism. CAC (call admission control) is one of QoS mechanisms and it maintains performance by limiting system utilization. However, the most conventional CAC algorithms are utilizatoin-based and measurement-based CAC algorithm, both have become important research topics. Among the research work, most of them wants to increase system utilization but not to degrade the quality of service. In this thesis, we propose a call admission control algorithm, which handles longlive and short-live dlows differently. Under such mechanism, we show that not only one can guarantee the quality of each flow, but it is also possible to improve the system utilization. Finally, we employ simulation to valide the effectiveness of this CAC algorithm to support streaming media services.Contents Abstract i 1 Introduction 1 1.1 A Brief Review of Call Admission Control Algorithms . . . . . . . . . 1 1.2 Long-lived and Short-lived Flows . . . . . . . . . . . . . . . . . . . . 3 2 System Models 5 2.1 Traffic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2 Queueing Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.3 The Setting of Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.4 Shaped Service Curves . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.4.1 Fast Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.4.2 Bursty Arrival . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3 Performance Analysis 15 3.1 Time-to-Play . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2 End-to-end Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1 The Maximum Tolerable Delay . . . . . . . . . . . . . . . . . 21 3.2.2 Delay Bounds of Long-lived Flows . . . . . . . . . . . . . . . 24 3.2.3 Delay Bounds of Short-lived Flows . . . . . . . . . . . . . . . 27 3.3 LS Call Admission Control Algorithm . . . . . . . . . . . . . . . . . . 29 4 Simulation Results of Call Admission Control Algorithms 33 4.1 Simulation Environments . . . . . . . . . . . . . . . . . . . . . . . . . 33 4.1.1 Parameters of Flows and Shapers . . . . . . . . . . . . . . . . 33 4.1.2 Traffic Pattern and the Arrival Pattern of Requests . . . . . . . . 34 iii iv 4.2 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.2.1 Time-to-Play . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 4.2.2 End-to-end Delay . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.2.3 System Utilization . . . . . . . . . . . . . . . . . . . . . . . . 39 5 Conclusions 43 5.1 Implementation Issues . . . . . . . . . . . . . . . . . . . . . . . . . . 43 5.2 Conclusions and Future Work . . . . . . . . . . . . . . . . . . . . . . . 44624449 bytesapplication/pdfen-US允入控制長短封包流Call Admission Controlthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/58624/1/ntu-93-R91942028-1.pdf