Scholarlink

<p>Autonomous data collection in the Internet of Underwater Things (IoUT) remarkably impacts coral monitoring, pollution monitoring, disaster monitoring, and military security. The IoUT has transformed resource management, but challenges like communication overhead, data privacy, and data heterogeneity persist. Integrate federated learning (FL), a decentralized machine learning with IoUT, to address IoUT challenges. FL performs decentralized learning on distributed data without data exchange to the server, enhancing the system’s efficiency and data privacy. This article inspects the benchmark parameters using a case study on coral classification and evaluates metrics to validate the performance of FL in IoUT. This article summarizes the components of IoUT and FL methods. This survey provides an outlook on the business aspect and the sustainability of the economic benefits for FL in IoUT. Besides that, this article discusses ethical and regulatory compliance to avoid misuse of FL in IoUT. The article also explores the advantages of FL in different applications and analyzes the challenges for future research.</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 learning-based privacy-preserving Internet of Underwater Things: a vision, architecture, computing, taxonomy, and future directions

  • 期刊论文

  • 发表日期：

    2025年5月21日

查看全文
  • Kamalika Bhattacharjya,
  • Debashis De

摘要

Autonomous data collection in the Internet of Underwater Things (IoUT) remarkably impacts coral monitoring, pollution monitoring, disaster monitoring, and military security. The IoUT has transformed resource management, but challenges like communication overhead, data privacy, and data heterogeneity persist. Integrate federated learning (FL), a decentralized machine learning with IoUT, to address IoUT challenges. FL performs decentralized learning on distributed data without data exchange to the server, enhancing the system’s efficiency and data privacy. This article inspects the benchmark parameters using a case study on coral classification and evaluates metrics to validate the performance of FL in IoUT. This article summarizes the components of IoUT and FL methods. This survey provides an outlook on the business aspect and the sustainability of the economic benefits for FL in IoUT. Besides that, this article discusses ethical and regulatory compliance to avoid misuse of FL in IoUT. The article also explores the advantages of FL in different applications and analyzes the challenges for future research.