I. Introduction
Real-time systems typically process and react to external events. In many complex real-time systems, the processing of the external event is performed by a processing chain consisting of multiple tasks which may either execute concurrently on the same processor or distributed to multiple processing units. A critical real-time performance metric of such systems is the maximum reaction time, which describes how long it takes from the event occurrence to when the final output corresponding to this external event is produced. The processing chains with maximum reaction time constraints can be widely found in many applications domains, e.g., automotive electronics, industry control and robotics.