Abstract

End-to-end throughput guarantee is an important service semantics that network providers would like to offer to their customers. A network provider can offer such service semantics by deploying a network where each router employs a fair packet scheduling algorithm to allocate network bandwidth to competing flows. Unfortunately, these scheduling algorithms require every router to maintain per-flow state and perform per-packet flow classification; these requirements limit the scalability of the routers. This paper makes two primary contributions: (1) We present the first tight analysis for deriving end-to-end throughput guarantees for a network of routers that employ per-flow fair scheduling algorithms. (2) We propose the Core-stateless Guaranteed Throughput (CSGT) network architecture that, without maintaining per-flow state or performing per-packet flow classification in core routers, provides to flows throughput guarantees that are within a constant of what is attained by a network of core-stateful fair routers. 1

Citations