CRITICALITY-AWARE PARTITIONED TASK SCHEDULING WITH LIMITED MIGRATORY ON MULTICORE PLATFORMS
Mixed-critical (MC) systems, in which different functionalities of varying criticality levels may consolidate on a shared embedded platform, are an active area of research in safety-related environments. With the proliferation of MC system, the multicore processor is becoming the obvious design choice in current and future safety-critical domains. The real-time scheduling of certifiable MC systems on a multicore platform has been recognized as a great challenging issue, where using conventional scheduling algorithms may cause significant under-utilization of the platform’s resources. In this work, we address this important dispute by proposing an effective optimal partitioning approach, the Criticality-aware Partitioned Algorithm (CaPA), that enables a limited number of migration of low-criticality workloads to improve the effectiveness of the schedulability by integrating the potential benefits of partitioned scheduling approaches. The results from extensive simulation under different situations demonstrate that CaPA always significantly outperforms existing MC partitioning heuristics in terms of acceptance ratios.
NAGALAKSHMI K AND GOMATHI N
To read the full article Download Full Article