Scholarlink

This research investigates the application of heuristic and reinforcement learning approaches to enhance resource efficiency in edge-based distributed systems. The proposed approaches aim to optimize resource allocation, workload balancing, resource provisioning, task scheduling, and average/maximum dilation metrics. Experimental results demonstrate significant improvements across all evaluation metrics. Resource allocation improved by 90–96%, workload balancing by 85–91%, resource provisioning by 92%–97%, and task scheduling by 90–96%. Average/maximum dilation values decreased from 0.05 to 0.04, indicating improved data flow management. These findings highlight the potential of heuristic and reinforcement learning techniques to optimize resource management in edge computing environments. The adaptive nature of these approaches enables them to dynamically adjust resource provisioning based on system demands, ensuring optimal utilization and responsiveness. In conclusion, research on improving resource efficiency with heuristic and reinforcement learning approaches in edge-based distributed systems has shown state-of-the-art promises. Thus, it renders the management of resources in an edge computing environment much brighter in terms of performance, efficiency, and responsiveness. Integration of these novel approaches will, therefore, allow attaining maximal benefit for distributed systems serving different applications throughout the evolution of edge computing.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Resource Efficiency Through Heuristic and Reinforcement Learning Approaches in Edge-Based Distributed Systems

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Namita Nath,
  • Neha Garg,
  • Jagendra Singh,
  • Ch. Lavanya,
  • Shikha Mittal,
  • Sumit Kumar,
  • Garima Jaiswal

摘要

This research investigates the application of heuristic and reinforcement learning approaches to enhance resource efficiency in edge-based distributed systems. The proposed approaches aim to optimize resource allocation, workload balancing, resource provisioning, task scheduling, and average/maximum dilation metrics. Experimental results demonstrate significant improvements across all evaluation metrics. Resource allocation improved by 90–96%, workload balancing by 85–91%, resource provisioning by 92%–97%, and task scheduling by 90–96%. Average/maximum dilation values decreased from 0.05 to 0.04, indicating improved data flow management. These findings highlight the potential of heuristic and reinforcement learning techniques to optimize resource management in edge computing environments. The adaptive nature of these approaches enables them to dynamically adjust resource provisioning based on system demands, ensuring optimal utilization and responsiveness. In conclusion, research on improving resource efficiency with heuristic and reinforcement learning approaches in edge-based distributed systems has shown state-of-the-art promises. Thus, it renders the management of resources in an edge computing environment much brighter in terms of performance, efficiency, and responsiveness. Integration of these novel approaches will, therefore, allow attaining maximal benefit for distributed systems serving different applications throughout the evolution of edge computing.