Hierarchical Distributed Distortion Aware Beamforming with Dynamic User Association & BS Clustering for Cell-Free Massive-MIMO Systems

Abstract

Massive multiple-input multiple-output (CF-mMIMO) systems have been identified as a promising technique for wireless communication, in terms of the enhancement of spectral efficiency, network coverage and user fairness. The key issues to be addressed for practical deployment are impairments in the hardware, inter-cell interference, high computational complexity, and high number of backhaul signalling, all of which impact system performance. To tackle these problems, in this paper, a Hierarchical Cluster-Based Distortion-Aware User Association (HC-DA-UA) framework for the distributed CF-mMIMO system under hardware impairments is proposed to solve these problems. The proposed method first clusters the distributed base stations based on the channel energy in a dynamic way to minimize the coordination overhead. Then a hierarchical distributed beamforming method is used to optimise the precoding in each cluster to minimise distortion caused by non-ideal hardware components. In addition, a distortion-aware user association mechanism is used to select the best serving base station for each user, taking into account the signal received and the distortion induced by the hardware and the interference. A second beamforming module called an interference estimation module further enhances the beamforming process to yield the most optimal achievable beamforming signal-to-interference-plus-noise-and-distortion ratio (SINDR). The simulation results reveal that the proposed HC-DA-UA framework always outperforms the existing counterparts in terms of sum-rate, whilst substantially decreasing the computational complexity and signalling overhead, thus it is an efficient and scalable solution for next-generation hardware-impaired distributed wireless communication networks.

Introduction

This paper proposes a Hierarchical Cluster-Based Distortion-Aware User Association (HC-DA-UA) framework for improving the performance of cell-free massive MIMO (CF-mMIMO) networks for future 5G/6G wireless communication.

1. Background and Problem

5G and 6G require high data rates, massive connectivity, low latency, and high spectral/energy efficiency. Conventional cellular networks suffer from:

• Inter-cell interference
• Limited coverage and user fairness
• Scalability problems
• High signaling and processing requirements

Massive MIMO improves capacity and spectral efficiency through many antennas, but centralized systems still face interference and scalability limitations. Cell-free massive MIMO, where geographically distributed base stations (BSs) jointly serve users without fixed cell boundaries, can improve coverage, fairness, and interference management. However, CF-mMIMO introduces hardware impairments, computational complexity, backhaul overhead, and difficult user association.

2. Research Gap

Previous studies have separately investigated technologies such as:

• Massive/XL-MIMO
• Hybrid beamforming
• Intelligent reflecting surfaces
• Distributed beamforming
• User-centric CF-mMIMO

However, fewer studies address dynamic clustering, hierarchical beamforming, hardware distortion, user association, computational complexity, and backhaul overhead together. The proposed HC-DA-UA framework aims to address these challenges in a unified approach.

3. System Model

The system consists of geographically distributed base stations jointly communicating with users. The model accounts for:

• Beamforming
• Multi-user interference
• Additive noise
• Nonlinear transmitter hardware distortion

Performance is evaluated using Signal-to-Interference-plus-Noise-and-Distortion Ratio (SINDR) and the resulting network sum rate:

Rsum?=k∑?log2?(1+γk?)

where γk? represents the effective SINDR of user k.

4. Proposed HC-DA-UA Method

The proposed framework has five main stages:

1. Dynamic Clustering – Distributed BSs are grouped dynamically according to channel conditions rather than using fixed clusters.
2. Hierarchical Beamforming – Beamforming is optimized locally within each cluster, reducing the need for network-wide optimization.
3. Distortion-Aware User Association – Users are associated with BSs based not only on channel quality but also on hardware distortion and beamforming performance.
4. Interference Estimation – Inter-cluster interference is estimated and incorporated into optimization to improve signal quality.
5. SINDR and Sum-Rate Evaluation – SINDR is calculated and the sum rate is optimized iteratively until convergence.

5. Results

MATLAB simulations compare HC-DA-UA with Ring, Star, Centralized, and conventional Distortion-Aware User Association (DA-UA) schemes.

The proposed method demonstrates:

• Higher sum rate across different transmit-power levels.
• Faster convergence, reaching high performance in fewer iterations.
• Improved scalability as the number of BSs increases.
• Higher throughput with increasing antenna numbers due to improved beamforming and spatial multiplexing.
• Lower computational complexity, because optimization is performed within dynamically formed clusters rather than across the entire network.
• Lower backhaul signaling overhead, because only necessary information is exchanged between clusters.

Conclusion

The paper introduced a Hierarchical Cluster-Based Distortion-Aware User Association (HC-DA-UA) for hardware-impaired cell-free massive MIMO (CFmnMIMO) system. The proposed framework combines dynamic clustering, hierarchical beamforming, distortion aware user association with interference estimation to enhance network performance with low computational complexity and backhaul overheads. The proposed method can effectively reduce the interference and hardware-induced distortion, improving the reliability and scalability of communication by performing optimized operations in a cooperative cluster at the local level. The simulation results using MATLAB confirmed that the HC-DA-UA framework consistently achieves higher sum rate when compared to conventional Ring, Star, Centralized and Distortion-Aware User Association (DA-UA) schemes in terms of computational effectiveness, signalling overhead and sum rate. Based on the obtained outcomes, the proposed framework is efficient and scalable solution for next generation distributed wireless communication network. In the future, the proposed HC-DA-UA will be expanded to other more practical and dynamic wireless communication scenarios. The framework can be improved with AI and RL techniques for time-varying environments with adaptive beamforming and user association. Also, the use of intelligent reflecting surfaces (IRS), reconfigurable intelligent surfaces (RIS), non orthogonal multiple access (NOMA) and XL-MIMO may be investigated to further enhance the spectral and energy efficiencies. The proposed framework can also be tested against user mobility, imperfect channel state information, heterogeneous network deployment, and hardware impairment models of realistic hardware impairments to ensure the robustness of 6G wireless communication systems.

References

[1] M. F. Haider, F. You, S. He, T. Rahkonen, and J. P. Aikio, “Predistortion-Based Linearization for 5G and beyond Millimeter-Wave Transceiver Systems: A Comprehensive Survey,” IEEE Communications Surveys and Tutorials, vol. 24, pp. 2029–2072, 2022, doi: 10.1109/COMST.2022.3199884. [2] S. Elhoushy, M. Ibrahim, and W. Hamouda, “Cell-Free Massive MIMO: A Survey,” IEEE Communications Surveys and Tutorials, vol. 24, pp. 492–523, 2022, doi: 10.1109/COMST.2021.3123267. [3] A. Vijay, S. Lakshmi Narayanan, N. K. Basha, K. Vijayakumar, K. Anusulatha, and G. S. Satheesh Kumar, “Beamforming Strategies in Massive MIMO: Evolution and Optimization Techniques,” in Proceedings of the 4th International Conference on Innovative Mechanisms for Industry Applications, ICIMIA 2025, Institute of Electrical and Electronics Engineers Inc., 2025, pp. 416–424. doi: 10.1109/ICIMIA67127.2025.11200524. [4] Abueidaet al., “Signal Processing for Cell-Free Massive MIMO: Techniques and Trends in Estimation, Detection, and Precoding,” 2025, Institute of Electrical and Electronics Engineers Inc.doi: 10.1109/OJCOMS.2025.3625823. [5] H. I. Obakhena, A. L. Imoize, F. I. Anyasi, and K. V. N. Kavitha, “Application of cell-free massive MIMO in 5G and beyond 5G wireless networks: a survey,” Dec. 2021, Springer Science and Business Media B.V.doi: 10.1186/s44147-021-00014-y. [6] Z. Wang et al., “A Tutorial on Extremely Large-Scale MIMO for 6G: Fundamentals, Signal Processing, and Applications,” IEEE Communications Surveys and Tutorials, vol. 26, pp. 1560–1605, 2024, doi: 10.1109/COMST.2023.3349276. [7] B. Zheng, C. You, W. Mei, and R. Zhang, “A Survey on Channel Estimation and Practical Passive Beamforming Design for Intelligent Reflecting Surface Aided Wireless Communications,” IEEE Communications Surveys and Tutorials, vol. 24, pp. 1035–1071, 2022, doi: 10.1109/COMST.2022.3155305. [8] N. Shlezinger, M. Ma, O. Lavi, N. T. Nguyen, Y. C. Eldar, and M. Juntti, “AI-Empowered Hybrid MIMO Beamforming,” Mar. 2023. [9] D. Khan et al., “A Holistic Review of Holographic MIMO for 6G: Hardware Design and Implementation Perspectives,” IEEE Open Journal of the Communications Society, vol. 7, pp. 4483–4524, 2026, doi: 10.1109/OJCOMS.2026.3686376.

Copyright

Copyright © 2026 Penchala Sri Poojitha, Dr. D. Karunakar Reddy. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.