Service Optimization of In-elastic Flows Stripped over Multiple Asymmetric Mobile Channels

摘要

In this paper, we present a stateless multi-server guaranteed rate scheduling service model that utilizes multiple, asymmetric wireless channels to maximize service guarantees for in-elastic flows. One of the major consequences of multi-channel flow forwarding is the increased intra-path and inter-path end-to-end delay variations between successive packets of a stream. These variations cause problems in achieving desired service levels due to increased reorder overhead, higher packet drop-rate, and higher buffer maintenance overhead. We have aggregated the capacity of available wireless channels through a link monitoring function (LMF) and has formulated a multi-server scheduling model that ensures tight delay-bounds and lower service latency to minimize inter-path delay variations and out-of-sequence arrival. The simulation results show consistently higher performance metrics of the scheme w.r.t end-to-end delay variations, in-order arrival, and service latency that enable reliable, scalable, guaranteed rate service during mobility.