ATM rate-based congestion control using a Smith predictor

This paper presents a feedback control algorithm for ATM congestion control in which ABR source rates are throttled according to VC queue levels at intermediate nodes along the path. The goal is to ''fill in'' the residual bandwidth, without exceeding a specified queue threshold. In order to obtain this, we propose a simple and classical proportional controller, plus a Smith Predictor to overcome instabilities due to large propagation delays. As a result, each queue behaves as a simple first-order dynamic system with a delay in cascade. The delay is out of the feedback loop, and therefore does not affect stabilitity. Moreover, since the system dynamic is a first-order one, it is not only stable but it does not even have damped oscillations. We show that this rate-based control scheme can actually be interpreted as a type of an end to end credit scheme. Finally, we propose an effective EPRCA implementation in which each source computes its input rate based on the maximum queue level along the path. Theoretical and experimental results show the fairness of the proposed control scheme, its efficiency under the constraints of the EPRCA implementation, as well as its cell loss free property. (C) 1997 Elsevier Science B.V.