| Call Admission Control (CAC) for Asynchronous Transfer Mode (ATM) networks under various traffic types is studied in this thesis. In order to evaluate the performance of different CAC schemes, a generalized traditional traffic generator and a self-similar traffic generator are implemented as process models in OPtimized Network Engineering Tool (OPNET). The generalized traditional traffic generator includes the deterministic ON/OFF, the two-state Markov Modulated Fluid Process (MMFP), the Interrupted Poisson Process (IPP), and the two-state Markov Modulated Poisson Process (MMPP) traffic models as special cases. The self-similar traffic generator is based on multiplexing a number of ON/OFF sources with heavy-tailed ON/OFF periods. Both traffic generators are capable of producing realistic ATM traffic at a very high speed of 25 millions cells per hour. Calls can be setup and released dynamically by those traffic generators.;A comprehensive literature review on existing CAC schemes has been carried out. The advantages and disadvantages of each of these schemes are discussed. A new CAC scheme based on estimating the available bandwidth of an output link is developed. It has no restrictions on the type of traffic sources and is able to operate in real-time. (Abstract shortened by UMI.). |
