Scholarlink

<p>Due to Internet of Things’ (IoT) rapid expansion, there are an unprecedented number of networked devices, which generates massive amounts of data and puts a significant energy strain on wireless sensor networks (WSNs). Traditional energy optimization methods are unable to handle the increasing level of complexity of IoT-integrated wireless network systems. The present research proposes the FDRL (N-federated deep reinforcement learning) Framework, an AI-driven deep learning system, to enhance energy efficiency (98%), enhance packet delivery ratio (96%), extend network lifetime (97%), and decrease latency extent up to 92%. The proposed framework uses reinforcement learning (RL) and federated learning (FL) to maximize utilization of resources while maintaining confidentiality of information. After federated averaging (data aggregation), centralized optimization is achieved using Pelican optimization (PO). The performance of the proposed framework is superior to the existing studies in context of IoT-WSN. The results show how deep learning and federated approaches can advance renewable energy optimization for future connected devices.</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

Federated deep reinforcement learning (FDRL) framework using pelican optimization (PO) to achieve sustainable energy in IoT-integrated wireless networks

  • 期刊论文

  • 发表日期：

    2025年10月24日

查看全文
  • Shivakumar Kagi,
  • Aruna M,
  • C. Gnana Kousalya,
  • R. John Martin,
  • Bhanu Sharma,
  • Vaibhav Prakash Vasani,
  • Sandip Nagpure

摘要

Due to Internet of Things’ (IoT) rapid expansion, there are an unprecedented number of networked devices, which generates massive amounts of data and puts a significant energy strain on wireless sensor networks (WSNs). Traditional energy optimization methods are unable to handle the increasing level of complexity of IoT-integrated wireless network systems. The present research proposes the FDRL (N-federated deep reinforcement learning) Framework, an AI-driven deep learning system, to enhance energy efficiency (98%), enhance packet delivery ratio (96%), extend network lifetime (97%), and decrease latency extent up to 92%. The proposed framework uses reinforcement learning (RL) and federated learning (FL) to maximize utilization of resources while maintaining confidentiality of information. After federated averaging (data aggregation), centralized optimization is achieved using Pelican optimization (PO). The performance of the proposed framework is superior to the existing studies in context of IoT-WSN. The results show how deep learning and federated approaches can advance renewable energy optimization for future connected devices.