Stochastic fluid models in the analysis of access regulation in high speed networks

The authors model analytic techniques for studying access regulation in asynchronous transfer mode (ATM)-based broadband integrated service digital networks (ISDNs). The traffic in these networks is expected to be highly diverse and bursty, and regulation will be required as an integral part of congestion control. They study a four-parameter regulator which is a generalized leaky bucket and which acts as a policing device as well as traffic shaper. The source and the regulator are described by a fluid model in which Markov modulated fluid sources reflect bursty characteristics. It is shown that with proper design the regulator effectively controls a three-way tradeoff between throughput, delay, and burstiness. Numerical results reveal that performance is sensitive to certain source characteristics. The regulator output, which consists of coupled streams of marked and unmarked cells, is characterized.<
>