Aco and Ga based fault-tolerant scheduling of real-time tasks on multiprocessor systems — A comparative study

Abstract

Fault-tolerant scheduling of real-time (RT) tasks in multiprocessor environment is essentially a NP-hard problem. This basically involves allocating a set of tasks to a set of processors so as to minimize the makespan and ensure tasks to meet their timing constraints. Many traditional heuristic approaches, such as earliest deadline first (EDF) and least laxity first (LLF) have been adopted to find optimal solution to this scheduling problem. However, conventional approach to achieve fault-tolerance (FT) in scheduling RT tasks based on traditional heuristic approach suffers from poor performance and results in inefficient processor utilization. Nature-inspired heuristic algorithms are gaining increased acceptance among researcher for solving real world NP-hard combinatorial optimization problems. This paper presents a comparative study of the novel primary-backup (PB) based fault-tolerant scheduling (PBFTS) technique for RT tasks in multiprocessor environment using two popular nature-inspired heuristic algorithms: the Ant Colony Optimization (ACO) and the Genetic Algorithm (GA). Exhaustive simulation reveals that the PBFTS algorithm based on GA and ACO both outperform the traditional PBFTS schemes in terms of performance, system utilization and efficiency. However, the comparative study also shows that the ACO based scheme surpasses the GA based scheme in terms of speed of execution whereas GA based scheme displays superior convergence with respect to ACO counterpart.