Improved EDF Algorithm for Fault Tolerance with Energy Minimization

The aim is to achieve schedulability of tasks in a real time system with fault tolerance and energy minimization. The fault tolerance can be achieved by maintaining enough time redundancy so that task can be re-executed in presence of fault. During re-execution, check pointing policy gives reliability in a system. First, calculate an optimal number of checkpoints to reduce the redundancy and save the system to complete re-execution. Energy minimization can be achieved by DVFS (Dynamic Voltage Frequency & Scaling). In this paper, existing non-preemptive EDF scheduling algorithm has been modified for fault tolerance and energy minimization. We adjust the voltage level according to available storage energy in the system & find feasibility test on each task. The worst case execution time is associated with voltage level. The approach is developed to enhance task schedulability and minimize energy consumption in presence of fault. At the end of the paper, experimental results shows that proposed algorithm is better than existing algorithm.