| The integration of data, voice, and video services in the emerging B-ISDN/ATM networks presents new challenges in traffic management. A clear understanding of the traffic characteristics of each of these services is necessary in order to guarantee the Quality of Service (QoS) to each traffic type. In this dissertation we develop a new descriptive model of traffic characterization process along with algorithms and tools to automate it. The use of sampling techniques and their effect on protocol distributions is studied and integrated as part of the traffic characterization process. In addition, we present several measured traffic model of Wide Area Network TCP/IP traffic collected at The junction of the Pennsylvania State University Data Backbone to the Internet. Finally, we apply the various empirical traffic modelling to solve some ATM traffic management problems such as queue and memory management, and connection admission control. |
