Evaluation Framework for Task Scheduling Algorithms in Distributed Reconfigurable Systems

Abstract

In this dissertation, we propose the design of a simulation framework to investigate the performance of reconfigurable processors in distributed systems. The framework incorporates the partial reconfigurable functionality of the reconfigurable nodes. Depending on the available reconfigurable area, each node is able to execute more than one task simultaneously. As part of implementation of the framework, we describe a simple mechanism for the resource information maintenance. We propose the design of data structures, which are essential parts of a Resource Information System (RIS). A detailed example is provided to discuss the basic functionality of these data structures, which maintain the information regarding the reconfigurable nodes, such as their updated statuses, their available areas, and the current tasks etc. Furthermore as a case study, we present a variety of scheduling strategies implemented to distribute tasks among reconfigurable processing nodes, utilizing the option of partial and full reconfigurability of the nodes. We propose a generic scheduling algorithm which is capable of assigning tasks to these two variants of the nodes. Using a given set of simulation parameters under the same simulation conditions, we performed various experiments. Based on the results, it is proved that the nodes with partial reconfigurable options provide a less average waiting time per task and total task completion time. In addition, the results suggest that the average wasted area per task is less as compared to the full configuration, verifying the functionality of the simulation framework.