Energy Minimization for Fault Tolerant Scheduling of Periodic Fixed-Priority Applications on Multiprocessor Platforms

An Efficient Fault-Tolerant Scheduling Approach with Energy Minimization for Hard Real-Time Embedded Systems

“…This algorithm combines DVFS for optimizing energy consumption and re-execution recovery for fault tolerance, but their scope is restricted to single processor systems. In [18], authors introduced an efficient method to determine the checkpointing scheme that can tolerate k transient faults on a single processor. They also proposed a task allocation scheme to reduce energy consumption.…”

“…The best response time WCRT best of the task can be written as: (18) WCRT best ( ) = min 1≤ ≤2 WCRT( ),…”

Energy-aware Fault-tolerant Scheduling for Hard Real-time Systems

“…As shown in [55], checkpointing with roll-back recovery is a very efficient technique to reduce recovery overhead and improve system schedulability. To our best knowledge, there is no work that targets on improving system schedulability for fixed-priority tasks on multi-core platforms based on exploring the combination of task partitioning and checkpointing.…”

“…Since there is no existing work in the literature that solves the exact same problem, we first extend the BFD to incorporate the checkpointing feature. The tasks are sorted in non-increasing order of utilization, the algorithm allocates one task at a time and use algorithm ECHK [55] to search for a feasible checkpointing scheme for this task and tasks on each candidate core. Among all feasible cores (a core is considered feasible if there exists a feasible checkpointing configuration among the tasks on the core and the task to be allocated), the task is assigned to the core with the least remanning utilization.…”

“…In particular, we consider that the system is subject to a maximum of K faults during one operation cycle of the system (its length is the least common multiple (LCM) of all task periods and is denoted by L). We adopt this K-fault model for the following three reasons: 1) it is a widely accepted fault model and well studied in the literature [124,57,54,130];…”