Time scale analysis of explicit rate allocation in ATM networks

Rate based congestion management for computer communication networks has been the focus of considerable research. We examine some of the fundamental properties of such end-to-end rate based algorithms. The first is the requirement for a policy to ensure the feasibility of the transmission rates by the sources. The feasibility implies that the capacity of any of the resources is not exceeded at any time. We show that a source policy which defers increases in the explicit rate for a period of time, while allowing decreases to take effect immediately maintains the requirement of feasibility. We also examine the convergence properties of the end-to-end rate based schemes in general, and give an upper bound on the convergence time of the scheme presented