In [18], we propose a framework to study a stream of trafic or message as it is transferred over an ATM connection. A message is modeled as a deierministic fluid flow, and an ATM node is modeled as a server which allocates bandwidth among concurrent messages. The key concept is that of the burstiness curve b(p) of an incoming message which gives the buffer size needed if the message is served at rate p. It is shown there that the fixed rate, the leaky bucket, and the affine server are burst reducing. This paper presents the bursliness bound for each of these servers. We also relate a sequence of identical afine servers to a fixed rate server, and suggest how ‘quality of service’ parameters may usefully be based on the burstiness curve.
Steven H. Low, Pravin Varaiya