GDM4MMIMO: Generative Diffusion Models for Massive MIMO Communications

作者：Zhenzhou Jin, You Li, H. Zhou, Yuanshuo Wang, Xiaofeng Liu, Xinrui Gong, Xiqi Gao, Derrick Wing Kwan Ng, Xiang‐Gen Xia · 发表于：IEEE Communications Magazine · 年份：2026 · DOI：10.1109/mcom.001.2500399 · 被引用次数：3 · 研究领域：Advanced MIMO Systems Optimization、Cooperative Communication and Network Coding、Advanced Wireless Communication Techniques

Massive multiple-input multiple-output (MIMO) offers significant advantages in spectral and energy efficiencies, positioning it as a cornerstone technology of fifth-generation (5G) wireless communication systems and a promising solution for the burgeoning data demands anticipated in sixth-generation (6G) networks. Meanwhile, the rapid evolution of artificial intelligence (AI) has ushered in a new era dominated by large AI models (LAMs), particularly large generative foundation models (LGFMs), which have achieved impressive success in computer vision (CV), natural language processing (NLP), and autonomous driving. As a pioneering force, these models are driving the paradigm shift in AI towards large generative AI (LaGenAI). Among them, the generative diffusion model (GDM), as one of state-of-theart families of generative models, demonstrates an exceptional capability to learn implicit prior knowledge and robust generalization capabilities, thereby enhancing its versatility and effectiveness across diverse applications. In this paper, we delve into the potential applications of GDM in massive MIMO communications. Specifically, we first provide an overview of massive MIMO communication, the framework of LGFMs, and the working mechanism of GDM. Following this, we discuss recent research advancements in the field and present a case study of near-field channel estimation based on GDM, demonstrating its promising potential for facilitating efficient ultra-dimensional channel stateme...