Scholarlink

<p>In large-scale Internet of Things (IoT) deployments using LoRaWAN, efficient gateway placement is critical for optimizing energy consumption, packet delivery, coverage, and latency. This paper presents a hybrid gateway placement algorithm (Hybrid-GPA) that integrates K-means clustering with Particle Swarm Optimization (PSO) for globally refined gateway selection. The approach first applies spatial clustering to identify candidate gateway locations, followed by PSO-based refinement to select an optimal subset that maximizes a composite utility function incorporating energy, latency, packet delivery ratio (PDR), and coverage. Additionally, utility-driven transmission power and spreading factor (SF) configurations are assigned to nodes under SINR constraints, ensuring link reliability. An iterative reconfiguration mechanism further adapts the network topology to dynamic conditions. MATLAB-based simulations demonstrate that Hybrid-GPA achieves a PDR of up to 92%, reduces average energy consumption to as low as 0.78&#xa0;mJ per node, and lowers end-to-end latency to as low as 190&#xa0;ms compared to state-of-the-art schemes such as BE-LoRa and ML Ensemble, while maintaining high coverage and scalability. These results validate the proposed framework as a robust and practical solution for intelligent gateway deployment in energy-constrained LoRaWAN-based IoT environments.</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

A hybrid optimization approach for gateway placement in IoT-enabled LoRaWAN using clustering and particle swarm optimization

  • 期刊论文

  • 发表日期：

    2026年5月7日

查看全文
  • Rashmi Kushwah

摘要

In large-scale Internet of Things (IoT) deployments using LoRaWAN, efficient gateway placement is critical for optimizing energy consumption, packet delivery, coverage, and latency. This paper presents a hybrid gateway placement algorithm (Hybrid-GPA) that integrates K-means clustering with Particle Swarm Optimization (PSO) for globally refined gateway selection. The approach first applies spatial clustering to identify candidate gateway locations, followed by PSO-based refinement to select an optimal subset that maximizes a composite utility function incorporating energy, latency, packet delivery ratio (PDR), and coverage. Additionally, utility-driven transmission power and spreading factor (SF) configurations are assigned to nodes under SINR constraints, ensuring link reliability. An iterative reconfiguration mechanism further adapts the network topology to dynamic conditions. MATLAB-based simulations demonstrate that Hybrid-GPA achieves a PDR of up to 92%, reduces average energy consumption to as low as 0.78 mJ per node, and lowers end-to-end latency to as low as 190 ms compared to state-of-the-art schemes such as BE-LoRa and ML Ensemble, while maintaining high coverage and scalability. These results validate the proposed framework as a robust and practical solution for intelligent gateway deployment in energy-constrained LoRaWAN-based IoT environments.