A simple approach to the cell loss probability in ATM networks

Estimating the cell loss probability in ATM networks is one of the most important problems concerning congestion control. A new approach to estimating the burst-level cell loss probability is presented. Our method is based on the approximation of the actual input process by means of a suitably chosen virtual two-state MMDP source, the parameters of which are determined by the burst-level cell loss probability without buffers and that with one buffer. A closed-form formula is obtained and simulation results show that our approach is sufficiently accurate for application.