Scholarlink

<p>In mobile edge computing, service migration promises intelligent task placement as well as maintaining seamless functionality in bandwidth-intensive, latency-sensitive applications. Despite the promise of service migration offered by deep reinforcement learning (DRL), existing techniques lack topology awareness and dynamic adaptability. This research proposes a novel graph convolutional network (GCN)-based DRL framework enhanced with multi-parameter feedback (energy, delay, bandwidth) for intelligent service migration in MEC. Unlike prior DRL methods, our GCN-based framework models server topologies as spatial relationships and adapts feedback weights dynamically in response to changing system conditions. Experimental evaluation using EdgeSimPy and EdgeAISim shows that our proposed method reduces power consumption by 27% compared to traditional DRL (DQN) and by 44% compared to heuristic approaches (Worst-Fit). Moreover, the framework consistently outperforms existing state-of-the-art DRL algorithms in terms of energy efficiency, latency reduction, and resource optimization. Our key contributions include: (1) A topology-aware GCN-DRL architecture; (2) Dynamic multi-parameter feedback-driven reward formulation that improve energy efficiency by 15% under high network stress; and (3) Comprehensive benchmarking against six algorithm classes in varied MEC scenarios.</p>

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

Optimizing edge intelligence: a DRL-driven service migration approach with enhanced feedback in mobile edge computing

  • 期刊论文

  • 发表日期：

    2025年7月17日

查看全文
  • Puneet Kansal,
  • Manoj Kumar,
  • O. P. Verma

摘要

In mobile edge computing, service migration promises intelligent task placement as well as maintaining seamless functionality in bandwidth-intensive, latency-sensitive applications. Despite the promise of service migration offered by deep reinforcement learning (DRL), existing techniques lack topology awareness and dynamic adaptability. This research proposes a novel graph convolutional network (GCN)-based DRL framework enhanced with multi-parameter feedback (energy, delay, bandwidth) for intelligent service migration in MEC. Unlike prior DRL methods, our GCN-based framework models server topologies as spatial relationships and adapts feedback weights dynamically in response to changing system conditions. Experimental evaluation using EdgeSimPy and EdgeAISim shows that our proposed method reduces power consumption by 27% compared to traditional DRL (DQN) and by 44% compared to heuristic approaches (Worst-Fit). Moreover, the framework consistently outperforms existing state-of-the-art DRL algorithms in terms of energy efficiency, latency reduction, and resource optimization. Our key contributions include: (1) A topology-aware GCN-DRL architecture; (2) Dynamic multi-parameter feedback-driven reward formulation that improve energy efficiency by 15% under high network stress; and (3) Comprehensive benchmarking against six algorithm classes in varied MEC scenarios.