An Experimental Analysis of Bottleneck-Centered Opportunistic Scheduling

Abstract

Bottleneck-centered, opportunistic scheduling is an effective problem solving paradigm based on a fundamental Analyze-Focus-Decide cycle. Opportunistic schedulers have shown better overall performance than other scheduling systems adopting different paradigms. However, in order to make more significant progress, we need to better understand how the different phases of the opportunistic cycle interact during problem solving. This paper reports an extensive analysis of a specific embodiment of the opportunistic scheduling paradigm, the Conflict Partition Scheduling procedure. By analyzing the role of each phase of CPS, proposing variations and anticipating expected effects on the performance of the algorithm, we aim to develop a better understanding of the importance of the different phases and of their interactions. The hypotheses on the performance of CPS are then verified in an experimental analysis. The results of this study provide hard evidence that supports the fundamental assumptions made by opportunistic scheduling.