Scholarlink

<p>Wireless Sensor Networks (WSNs) serve as the data sinks for many Internet of Things (IoT) applications, utilizing resource-constrained sensor nodes to monitor environmental conditions continuously. Despite the limited battery capacity and harsh deployment conditions, energy efficiency remains one of the most significant challenges in these networks. To alleviate this, we introduce two protocols that utilize reinforcement learning: Learning-Based Energy-Efficient Routing (LbEER) and Learning-Based Routing with Energy Balancing (LbREB). These protocols utilize AI methods to dynamically select routes based on real-time changes in node topology, energy, and communication conditions. MATLAB simulation results indicate that the performance of LbEER and LbREB is superior to that of benchmark protocols (LEACH, PEGASIS, and FlatEER-RL). In particular, LbEER reduces the energy consumption by up to 26% and increases the network lifetime by up to 48% compared to FlatEER-RL. However, for the dynamic condition, the improvement in the packet delivery ratio concerning LbREB is approximately 11%. While both protocols keep more than 40 active nodes at 4500 communication rounds, FlatEER-RL drops below 15 nodes. These results show the scalability, efficiency, and robustness of our approach. The protocols are well-suited solutions for energy-constrained and dynamic IoT environments, addressing energy usage balancing and enabling intelligent decision-making.</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

Implementation of novel learning based energy efficient routing protocols in wireless sensor networks for internet of things use cases

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年9月9日

查看全文
  • Kukunoor Shekar,
  • Nalavala Ramanjaneya Reddy,
  • Siddapuram Arvind,
  • T. V. Suresh Kumar,
  • Subrahmanyam Kodukula,
  • Geddam Varahagiri

摘要

Wireless Sensor Networks (WSNs) serve as the data sinks for many Internet of Things (IoT) applications, utilizing resource-constrained sensor nodes to monitor environmental conditions continuously. Despite the limited battery capacity and harsh deployment conditions, energy efficiency remains one of the most significant challenges in these networks. To alleviate this, we introduce two protocols that utilize reinforcement learning: Learning-Based Energy-Efficient Routing (LbEER) and Learning-Based Routing with Energy Balancing (LbREB). These protocols utilize AI methods to dynamically select routes based on real-time changes in node topology, energy, and communication conditions. MATLAB simulation results indicate that the performance of LbEER and LbREB is superior to that of benchmark protocols (LEACH, PEGASIS, and FlatEER-RL). In particular, LbEER reduces the energy consumption by up to 26% and increases the network lifetime by up to 48% compared to FlatEER-RL. However, for the dynamic condition, the improvement in the packet delivery ratio concerning LbREB is approximately 11%. While both protocols keep more than 40 active nodes at 4500 communication rounds, FlatEER-RL drops below 15 nodes. These results show the scalability, efficiency, and robustness of our approach. The protocols are well-suited solutions for energy-constrained and dynamic IoT environments, addressing energy usage balancing and enabling intelligent decision-making.