Degree Name

Master of Engineering


School of Electrical, Computer and Telecommunications Engineering


One of the challenges in ATM traffic management design is to maintain the quality of service for various classes while attempting to make maximum use of network resources. Available bit rate (ABR) mechanisms allow the network to divide the available bandwidth fairly and efficiently amongst the active sources. One of the difficulties in providing optimal control for ABR traffic is that the available bandwidth fluctuates dynamically due to statistical variations in high priory traffic and changes in the number of active ABR connections sharing the available bandwidth. It is apparent from the literature that reactive feedback control may not operate effectively in networks with a large delay bandwidth product. Therefore, the focus of this thesis is on development and evaluation of flow control algorithms for ABR in networks with a large delay bandwidth product. To evaluate the performance of the flow control algorithm, we have adopted both the analytical and simulation models. By matching the independent results obtained from the simulation and analytical approaches we can validate the accuracy of both models. The detailed simulation model utilizes bursty ABR traffic sources, realistic variable bit rate traffic traces, a finite switch buffer and retransmission protocol to describe the ATM network.