Scholarlink

In the evolving field of federated learning (FL), particularly within the internet of medical things (IoMT), the role of gradient compression emerges as a crucial element for enhancing communication efficiency and reducing associated costs. This literature review delves into the nuanced application of security measures integrated within gradient compression techniques in FL, especially in the context of IoMT. Through an exhaustive analysis of 14 significant conference papers and journal articles, the study brings to the forefront the various methodologies like quantization, sparsification, and adaptive algorithms. These methods are not only pivotal in addressing communication efficiency but also in managing operational costs in FL and IoMT applications. Despite extensive research in compression for FL, a notable gap persists in literature addressing the integration of these techniques with security protocols in IoMT communications. This chapter provides a much needed survey on this topic underscoring the need for dedicated research in this direction. As FL continues to revolutionize the field of IoMT, the integration of gradient compression with robust security protocols becomes imperative.

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 Review of Secure Gradient Compression Techniques for Federated Learning in the Internet of Medical Things

  • 章节

  • 发表日期：

    2025年

查看全文
  • Timoteo Kelly,
  • Ahmad Alhonainy,
  • Praveen Rao

摘要

In the evolving field of federated learning (FL), particularly within the internet of medical things (IoMT), the role of gradient compression emerges as a crucial element for enhancing communication efficiency and reducing associated costs. This literature review delves into the nuanced application of security measures integrated within gradient compression techniques in FL, especially in the context of IoMT. Through an exhaustive analysis of 14 significant conference papers and journal articles, the study brings to the forefront the various methodologies like quantization, sparsification, and adaptive algorithms. These methods are not only pivotal in addressing communication efficiency but also in managing operational costs in FL and IoMT applications. Despite extensive research in compression for FL, a notable gap persists in literature addressing the integration of these techniques with security protocols in IoMT communications. This chapter provides a much needed survey on this topic underscoring the need for dedicated research in this direction. As FL continues to revolutionize the field of IoMT, the integration of gradient compression with robust security protocols becomes imperative.