Scholarlink

<p>Efficient task scheduling in the Internet of Vehicles (IoV) is crucial for optimizing communication and computational resources, especially under stringent latency and reliability constraints. Traditional task scheduling methods often struggle to handle the complex interactions between vehicle-side energy consumption, edge-side operational costs, and system stability. To address this, we propose a novel two-layer hybrid framework that integrates the Improved Whale Optimization Algorithm (IWOA) with fractional programming (FP) and Lyapunov Drift-Plus-Penalty (DPP) for IoV task scheduling. Our approach decouples the global search of discrete decisions from the optimization of continuous variables, ensuring both efficiency and stability. Experimental results show that our method outperforms benchmark algorithms in terms of energy consumption and latency, achieving up to a 27.08% reduction in total energy consumption and a 25.56% improvement in average latency compared to existing solutions.</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

An improved whale optimization algorithm for efficient task scheduling in the internet of vehicles

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月28日

查看全文
  • Huiyong Li,
  • Shuhe Han,
  • Yang Yang

摘要

Efficient task scheduling in the Internet of Vehicles (IoV) is crucial for optimizing communication and computational resources, especially under stringent latency and reliability constraints. Traditional task scheduling methods often struggle to handle the complex interactions between vehicle-side energy consumption, edge-side operational costs, and system stability. To address this, we propose a novel two-layer hybrid framework that integrates the Improved Whale Optimization Algorithm (IWOA) with fractional programming (FP) and Lyapunov Drift-Plus-Penalty (DPP) for IoV task scheduling. Our approach decouples the global search of discrete decisions from the optimization of continuous variables, ensuring both efficiency and stability. Experimental results show that our method outperforms benchmark algorithms in terms of energy consumption and latency, achieving up to a 27.08% reduction in total energy consumption and a 25.56% improvement in average latency compared to existing solutions.