Temporal Partitioning on Multicore Platform

Paper i proceeding, 2014

This paper addresses the problem of ensuring temporal partitioning according to the ARINC-653 standard for integrating multiple applications on the same multicore platform. To employ temporal partitioning, we propose the design and analysis of a hierarchical scheduling framework (HSF) for multicore platform. In HSF, each application has a server task, which is mapped to one of the physical cores of the multicore platform. The HSF framework is based on scheduling at two-levels: (i) a system-level scheduler for each core schedules the server tasks that are mapped to that core, and (ii) a task-level scheduler for each application schedules the tasks of the application. This paper presents the design and analysis of this two-level HSF that can be used to ensure temporal partitioning and meeting all the deadlines of each application tasks. The effectiveness of our technique is demonstrated using real-world space applications provided by RUAG Space Sweden AB.