Techniques for adaptive estimation of effective bandwidth in ATM networks

This paper examines the effectiveness of adaptive algorithms for estimation of effective bandwidth in ATM networks. Large deviations results for constant capacity single server queues are used to derive a basic effective bandwidth result which is applied to a linear representation of the input process giving the effective bandwidth as a function of the cumulant generating function of the innovations, QoS, mean arrival rate and buffer size. The algorithm is tested by simulation using the least squares lattice algorithm for estimating the linear process coefficients for a range of input processes. The effectiveness of blocking and truncation of the input process on the algorithm is investigated.