An Evolutionary Multi-Task Optimization Algorithm Based on Affine Transformation in Edge Computing

In the domain of edge computing, one of the most significant challenges is the efficient allocation and scheduling of resources for multiple concurrent tasks given the constraints of limited computational resources. To address the resource optimization and scheduling challenges in edge computing, th...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Chinese Control Conference s. 2194 - 2200
Hlavní autori: Zhang, Zhe, Cao, Jianfu, Wang, Lin
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: Technical Committee on Control Theory, Chinese Association of Automation 28.07.2025
Predmet:
auxiliary affine transformation
Computational efficiency
Edge computing
Evolutionary multi-task optimization
Heuristic algorithms
Knowledge transfer
multi-factor evolutionary algorithm
Multitasking
Optimization
Processor scheduling
Resource management
resource optimization
Scheduling
task scheduling
Training
ISSN:1934-1768
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:In the domain of edge computing, one of the most significant challenges is the efficient allocation and scheduling of resources for multiple concurrent tasks given the constraints of limited computational resources. To address the resource optimization and scheduling challenges in edge computing, this paper introduces an Auxiliary Affine Transformation (AAT) algorithm designed to enhance cross-task alignment within the Evolutionary Multi-task Optimization (EMTO) framework. Specifically, AAT employs an iterative process where training samples are redistributed via affine transformation in each iteration. The algorithm dynamically establishes and optimizes task correspondences based on updated mapping results. By identifying the optimal alignment from existing transformations and iteratively refining the transformation matrix, AAT maximizes the similarity between heterogeneous tasks, thereby improving knowledge transfer efficiency in multi-task EMTO applications. Integrating AAT with EMTO addresses practical challenges in multi-task resource optimization and scheduling in edge computing environments. Experimental results show that the proposed method not only significantly enhances computational efficiency in task scheduling and optimization but also surpasses traditional methods in resource utilization and overall optimization effectiveness. In summary, AAT provides a more effective solution for resource optimization and task scheduling in edge computing while reinforcing the EMTO framework.
ISSN:1934-1768
DOI:10.23919/CCC64809.2025.11178867