A Practical Comparison of Scheduling Algorithms for Mixed Criticality Embedded Systems

Abstract: With the consolidation of automotive control processes onto single highperformance ECUs the issue of running, and thus scheduling, processes of varying criticality on a single CPU has moved to the fore. This has resulted in a number of new algorithms for scheduling such systems, for example Adaptive Mixed Criticality (AMC). This project attempts to measure the performance of some of these algorithms on a singlecore embedded system CPU and compares them in order to shed some light on their different advantages and disadvantages.