Fault-Tolerant Task Scheduling in Multicore Systems Using Hybrid Metaheuristic Algorithms
TL;DR
The experimental findings indicate that the combined strategy achieves shorter execution times, better scalability as workload increases, and a noticeable reduction in task failure rates, suggesting that integrating evolutionary and swarm-based optimization mechanisms can provide a practical and robust solution for improving both performance and fault tolerance for modern multicore systems.
Abstract
Ensuring reliable and efficient task scheduling remains a critical challenge in multicore computing environments, particularly when system faults can significantly affect performance and interfere with execution. This paper presents a hybrid optimization strategy that combines Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) techniques to improve task allocation under fault-prone conditions. The proposed model considers task dependencies during scheduling and dynamically distributes workloads across available processing cores to achieve balanced utilization while maintaining reliability.To evaluate its effectiveness, the hybrid GA–PSO method was tested against standalone GA and PSO approaches. The experimental findings indicate that the combined strategy achieves shorter execution times, better scalability as workload increases, and a noticeable reduction in task failure rates. These results suggest that integrating evolutionary and swarm-based optimization mechanisms can provide a practical and robust solution for improving both performance and fault tolerance for modern multicore systems.