A Unified Distributed Algorithm for Hybrid Near-Far Field Activity Detection in Cell-Free Massive MIMO

Jingreng Lei, Yang Li, Ziyue Wang, Qingfeng Lin, Ya‐Feng Liu, Yik‐Chung WuarXiv 2025 · DOI · arXiv

Abstract

A great amount of endeavor has recently been devoted to activity detection for massive machine-type communications in cell-free multiple-input multiple-output (MIMO) systems. However, as the number of antennas at the access points (APs) increases, the Rayleigh distance that separates the near-field and far-field regions also expands, rendering the conventional assumption of far-field propagation alone impractical. To address this challenge, this paper establishes a covariance-based formulation that can effectively capture the statistical property of hybrid near-far field channels. Based on this formulation, we theoretically reveal that increasing the proportion of near-field channels enhances the detection performance. Furthermore, we propose a distributed algorithm, where each AP performs local activity detection and only exchanges the detection results to the central processing unit, thus significantly reducing the computational complexity and the communication overhead. Not only with convergence guarantee, the proposed algorithm is unified in the sense that it can handle single-cell or cell-free systems with either near-field or far-field devices as special cases. Simulation results validate the theoretical analyses and demonstrate the superior performance of the proposed approach compared with existing methods.

Introduction

In the push toward 6G and massive Machine-Type Communications (mMTC), thousands of IoT devices require connectivity with minimal latency. "Grant-free" access is a promising solution where devices transmit data without prior permission. However, this necessitates that Access Points (APs) accurately perform Activity Detection to identify active devices.

Traditional algorithms assume "far-field" propagation, where signals arrive as plane waves. As AP antenna counts increase (Massive MIMO), the Rayleigh distance expands, creating a "hybrid" environment. In cell-free networks, a single device may be in the near-field (spherical waves) of one AP but the far-field of another. Standard covariance-based methods fail here as they assume independent, identically distributed (i.i.d.) signals, which does not hold for correlated near-field channels.

Methodology

We propose a covariance-based formulation to capture the unique statistical properties of hybrid channels. To solve the resulting optimization problem efficiently, we developed a two-tier approach:

  1. Distributed Framework (ADMM): Utilizing the Alternating Direction Method of Multipliers, the global detection problem is decomposed into parallel subproblems. APs perform local detection and only exchange activity estimates ($\theta_m$) with a Central Processing Unit (CPU), drastically reducing communication overhead compared to sending raw signals.
  2. Local Solver (Coordinate Descent with Taylor Expansion): We developed a novel Coordinate Descent (CD) algorithm using Sherman-Morrison-Woodbury updates and a first-order Taylor expansion. This allows APs to handle complex matrix inversions efficiently without sacrificing accuracy.

Theoretical Insight: The Near-Field Advantage

A major theoretical contribution of this work is the proof that near-field propagation is a feature, not a bug. We demonstrate that increasing the proportion of near-field channels actually enhances detection performance. This is due to the increased orthogonality of near-field channel vectors, which reduces cosine similarity and makes active devices easier to distinguish.

Theoretical logic showing how near-field devices improve detection performance through enhanced orthogonality.

Performance Evaluation

Simulation results confirm that the proposed algorithm consistently achieves lower error probabilities (Missed Detection and False Alarm) than benchmarks like Mismatched CD, Vanilla MLE, AMP, and OAMP. Notably, the distributed algorithm reaches the performance of an ideal centralized approach in just 2 iterations and provides a 14.7x speedup in a 12-AP setup (7.82s vs 114.96s).

Probability of error vs. total number of antennas, demonstrating the proposed method's superior accuracy over existing benchmarks.

Conclusion

This paper provides a unified, distributed framework for activity detection in hybrid near-far field cell-free MIMO systems. By leveraging the inherent benefits of near-field propagation and employing an efficient ADMM-based solver, we achieve state-of-the-art accuracy with significantly reduced computational and communication costs. Future work will explore robustness against imperfect synchronization and dynamic channel statistics estimation.

Summary updated

Abstract

A great amount of endeavor has recently been devoted to activity detection for massive machine-type communications in cell-free multiple-input multiple-output (MIMO) systems. However, as the number of antennas at the access points (APs) increases, the Rayleigh distance that separates the near-field and far-field regions also expands, rendering the conventional assumption of far-field propagation alone impractical. To address this challenge, this paper establishes a covariance-based formulation that can effectively capture the statistical property of hybrid near-far field channels. Based on this formulation, we theoretically reveal that increasing the proportion of near-field channels enhances the detection performance. Furthermore, we propose a distributed algorithm, where each AP performs local activity detection and only exchanges the detection results to the central processing unit, thus significantly reducing the computational complexity and the communication overhead. Not only with convergence guarantee, the proposed algorithm is unified in the sense that it can handle single-cell or cell-free systems with either near-field or far-field devices as special cases. Simulation results validate the theoretical analyses and demonstrate the superior performance of the proposed approach compared with existing methods.
BibTeX
@misc{lei2025unified,
  author = {Jingreng Lei and Yang Li and Ziyue Wang and Qingfeng Lin and Ya‐Feng Liu and Yik‐Chung Wu},
  title = {{A Unified Distributed Algorithm for Hybrid Near-Far Field Activity Detection in Cell-Free Massive MIMO}},
  year = {2025},
  eprint = {2509.15162},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2509.15162},
  howpublished = {arXiv},
}
Figure