Analytical Study on Scheduling Algorithms of Periodic Tasks in Real Time Systems on Multicore Processor

Real Time Systems are considered nowadays as the most common systems, because its wide spread and usage in many areas including technical and applied researches, also the installation of such systems on multicore platforms made them very desirable as embedded systems and control units, because of its high performance and robustness comparing to multiprocessors platforms. Scheduling is the basic operation in real time systems, and relies in ordering the execution of tasks depending on priorities witch is set according to scheduling policies. This paper aims to introduce an analytical study of the most important scheduling algorithms to find the one with best performance according to number of parameters such as system load, context switching overheads, and scheduling overheads, when they applied to number of periodic tasks generated randomly. We use for this purpose SIMSO real time systems simulator, because of its reliability, robustness, and the support to a large number of scheduling algorithms, and cache memory simulation with its different levels witch is considered to be the main component in multicore platforms.