BANDWIDTH SCHEDULING FOR WIDE-AREA ATM NETWORKS USING VIRTUAL FINISHING TIMES

This paper is concerned with the design of a class of bandwidth schedu ling policies that are suitable for public, wide-area asynchronous tra nsfer mode (ATM) networks. We specify design goals for such strategies including ease of implementation and the ability to guarantee minimum bandwidths to individual buffers. Packetized generalized processor sh aring is briefly discussed and a minimum bandwidth result for self-clo cked fair queueing is given. We revisit an approach originally propose d by Zhang and prove that it is appropriate for ATM. Some novel, relat ed approaches are described and analyzed.