Real-Time Tasks Scheduling with Value Control to Predict Timing Faults During Overload

Modern real-time applications are very dynamic and cannot cope with the use of worst case execution time to avoid overload situations. Therefore scheduling algorithms that are able to prevent timing faults during overload are required. In this context, the value parameter has become useful to add generality and flexibility to such systems. In this paper, we present the scheduling algorithm called DMB (dynamic misses based), which is capable of dynamically changing tasks value in order to adjust their importance according to its timing faults rate. The main goal of DMB is to allow the prediction of timing faults during overloads and thereby support a dynamic tuning of tasks fault rate. It is used to enhance the features of the previously defined TAFT (time-aware fault-tolerant) scheduler. Obtained results show that DMB in conjunction with TAFT reached the most promising results during overloads, allowing to control tasks degradation in a graceful and determined way