.png){kind=icon}
.png){kind=icon}
Return
Optimizing multiprocessor performance in real-time systems using an innovative genetic algorithm approach
.png){kind=icon}
OA
AI
DOI：10.1038/s41598-024-80910-4.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Due to its enormous influence on system functionality, researchers are presently looking into the issue of task scheduling on multiprocessors. Establishing the most advantageous schedules is often regarded as a difficult-to-compute issue. Genetic Algorithm is a recent tool employed by researchers to optimize scheduling tasks and boost performance, although this field of research is yet mostly unexplored. In this article, a novel approach for generating task schedules for real-time systems utilizing a Genetic Algorithm is proposed. The approach seeks to design task schedules for multiprocessor systems with optimal or suboptimal lengths, with the ultimate goal of achieving high performance. This research project focuses on non-preemptive independent tasks in a multiprocessor environment. All processors are assumed to be identical. We conducted a thorough analysis of the proposed approach and pitted it against three frequently utilized scheduling methodologies: the Evolutionary Fuzzy Based Scheduling Algorithm, the Least Laxity First Algorithm, and the Earliest Deadline First Algorithm. The Proposed Algorithm demonstrated superior efficiency and reliability compared to Earliest Deadline First, Least Laxity First, and Evolutionary Fuzzy-based Scheduling Algorithm. It consistently achieved zero missed deadlines and the lowest average response and turnaround times across all scenarios, maintaining optimal performance even under high load conditions.
Keywords:
Multiprocessors
Task Scheduling
Genetic algorithms
Performance utilization
Multiprocessor
No-Preemptions
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
3.9
Papers:
27.8W
Citations:
83.5W