| Summary: | 碩士 === 國立中央大學 === 統計研究所 === 95 === Switched Processing Systems (SPS) have received a lot of attentions in the past years. Over the last decade, research has been focused on how to obtain the maximal throughput of the system by constructing optimal scheduling policies which are executed throughout switching service modes (for example, MaxProduct). But, if the switch time of severs exchanging is non-negligible (theoretically, zero), this kind of policies will lose a lot of throughput. In 2003, Armony ＆ Bambos brought up Adaptive Batching (BatchAdapt) policies. No matter how long the switch time is, it can possess maximal throughput property. But the delay performance of BatchAdapt policies is very worse. So we utilize the MaxProduct policies to construct new cone policies with dynamical decision cones, called Expanded Cone policies. Finally, we also perform the simulation study to show that policies that we propose can almost achieve maximal throughput when the switch time is considered, and perform better than BatchAdapt policies in average delay.
|
|---|
