| 彭永倩,刘鸿.基于雾计算的分布式网络优化方法[J].高技术通讯(中文),2026,36(6):602~610 |
| 基于雾计算的分布式网络优化方法 |
| A distributed network optimization approach based on fog computing |
| |
| DOI：10. 3772 / j. issn. 1002 - 0470. 2026. 06. 006 |
| 中文关键词: 雾计算；自适应惯性权重；局部搜索；遗传交叉；约束惩罚机制；改进粒子群算法 |
| 英文关键词: fog computing, adaptive inertia weights, localized search, genetic crossover, constrained penalty mechanism, improved particle swarm algorithm |
| 基金项目: |
| 作者 | 单位 | | 彭永倩 | (黔南民族职业技术学院大数据与电子商务系都匀 558022) | | 刘鸿 | |
|
| 摘要点击次数: 71 |
| 全文下载次数: 69 |
| 中文摘要: |
| 针对智慧茶园环境中分布式网络面临的高延迟与低能效问题，本文提出一种基于雾计算的分布式网络优化算法。本文建立了综合考虑计算能力、通信带宽与能量消耗的多层级网络模型，提出了一个优化框架，能够高效调度任务并进行计算迁移。在此基础上，设计了融合自适应惯性权重、局部搜索、遗传交叉和约束惩罚机制的混合粒子群优化(hybrid particle swarm optimization，Hybrid-PSO)算法，该算法显著提升了全局搜索能力与解的可行性。实验结果表明，Hybrid-PSO算法在多目标优化问题上，能够有效减少系统总延迟、能耗和任务完成时间，收敛速度也显著优于传统的粒子群优化（particle swarm optimization，PSO）算法、遗传算法(genetic algorithm，GA)和PSO-GA算法。进一步分析表明，计算迁移与能量感知路由策略的引入，不仅优化了负载均衡，还显著延长了网络的寿命。 |
| 英文摘要: |
| This paper addresses the high latency and low energy efficiency issues in distributed networks within smart tea garden environments by proposing an optimization algorithm based on fog computing. A multi-layered network model is developed, considering computational capacity, communication bandwidth, and energy consumption. This model serves as the foundation for an optimization framework that efficiently allocates tasks and manages computational migration. The paper also introduces an hybrid particle swarm optimization (Hybrid-PSO) algorithm, which combines adaptive inertia weights, local search, genetic crossover, and a constraint penalty mechanism. This approach significantly enhances the algorithm’s global search ability and solution feasibility. Experimental results show that Hybrid-PSO reduces system latency, energy consumption, and task completion time, outperforming traditional particle swarm optimization (PSO), genetic algorithm (GA), and PSO-GA algorithms in terms of convergence speed. Further analysis indicates that computational migration and energy-aware routing strategies improve load balancing and extend the network’s lifetime. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
