Scheduling hard real-time tasks on multi-core using intelligent rate-monotonic

Abstract

Recently, researchers have applied semi-partitioned approaches to improve performance of hard real-time scheduling algorithms in multiprocessor architectures. RMLS is one of these methods. However, advantages of using semi-partitioned methods are often limited by well-known scheduling algorithms such as RM and EDF, which the former is simple but inefficient and the latter is efficient but has high processing overhead. There is an intelligent algorithm working on uniprocessor named IRM which takes advantages of both RM and EDF algorithms using that we present a new method called intelligent rate-monotonic least splitting to improve RMLS. Experimental results show that the proposed algorithm outperforms many other algorithms in literature in terms of processor utilization.