Device Activity Detection and Channel Estimation Using Score-Based Generative Models in Massive MIMO

doi:10.12142/ZTECOM.202501007

ZTE Communications ›› 2025, Vol. 23 ›› Issue (1): 53-62.DOI: 10.12142/ZTECOM.202501007

• Special Topic • Previous Articles Next Articles

Device Activity Detection and Channel Estimation Using Score-Based Generative Models in Massive MIMO

TANG Chenyue¹(), LI Zeshen¹, CHEN Zihan², YANG Howard H.¹

^1.ZJU-UIUC Institute, Zhejiang University, Haining 314400, China
^2.Singapore University of Technology and Design, Singapore 487372, Singapore

Received:2025-01-02 Online:2025-03-25 Published:2025-03-25
About author:TANG Chenyue (chenyue.23@intl.zju.edu.cn) received her BE degree from Central South University, China in 2023. She is currently working toward her master's degree in electrical engineering and information technology at Zhejiang University, China. Her current research interests include statistical signal processing, inverse problems, and machine learning.
LI Zeshen received his BE degree from Jilin University, China in 2022. He is working toward his PhD degree in information and communication engineering at Zhejiang University, China. His current research interests include federated learning, edge computing, and distributed machine learning.
CHEN Zihan received his BE degree in communication engineering from the Yingcai Honors College, University of Electronic Science and Technology of China (UESTC) in 2018 and PhD degree from the Singapore University of Technology and Design (SUTD)-National University of Singapore (NUS) Joint PhD Program in 2022. Currently, he is a postdoctoral research fellow with SUTD. His research mainly focuses on network intelligence, machine learning, and edge computing.
Howard H. YANG received his BE degree in communication engineering from Harbin Institute of Technology, China in 2012, and MSc degree in electronic engineering from the Hong Kong University of Science and Technology, China in 2013, and PhD degree in electrical engineering from the Singapore University of Technology and Design, Singapore in 2017. He is currently an assistant professor with Zhejiang University/University of Illinois at Urbana-Champaign Institute, Zhejiang University, China. His research interests currently focus on the modeling of modern wireless networks, high dimensional statistics, graph signal processing, and machine learning.
Supported by:
TANG Chenyue and LI Zeshen are co-first authors

Abstract

Abstract:

The growing demand for wireless connectivity has made massive multiple-input multiple-output (MIMO) a cornerstone of modern communication systems. To optimize network performance and resource allocation, an efficient and robust approach is joint device activity detection and channel estimation. In this paper, we present an approach utilizing score-based generative models to address the under-determined nature of channel estimation, which is data-driven and well-suited for the complex and dynamic environment of massive MIMO systems. Our experimental results, based on a comprehensive dataset generated through Monte-Carlo sampling, demonstrate the high precision of our channel estimation approach, with errors reduced to as low as -45 dB, and exceptional accuracy in detecting active devices.

Key words: activity detection, channel estimation, inverse problem, score-based generative model, massive MIMO

TANG Chenyue, LI Zeshen, CHEN Zihan, YANG Howard H.. Device Activity Detection and Channel Estimation Using Score-Based Generative Models in Massive MIMO[J]. ZTE Communications, 2025, 23(1): 53-62.

Figures/Tables 6

Figure 1 A step-by-step process for estimating X by employing a score-based model in conjunction with the known matrices Y and P

Figure 2 System model of a massive device communication network

Figure 3 An elaborate schematic representation of model sθ utilizing the RefineNet architecture. This fundamental block is cascaded D times in sequence

Figure 4 Training loss of the score-based generative model over steps

Figure 5 Performance comparison of channel estimation methods

Figure 6 Diagrams for simulaion results

References 44

1	CHEN X M, NG D W K, YU W, et al. Massive access for 5G and beyond [J]. IEEE journal on selected areas in communications, 2021, 39(3): 615-637. DOI: 10.1109/JSAC.2020.3019724
2	MIUCCIO L, PANNO D, RIOLO S. An energy-efficient DL-aided massive multiple access scheme for IoT scenarios in beyond 5G networks [J]. IEEE Internet of Things journal, 2023, 10(9): 7936–7959. DOI: 10.1109/JIOT.2022.3231128
3	VISWANATHAN H, MOGENSEN P E. Communications in the 6G era [J]. IEEE access, 2020, 8: 57063–57074. DOI: 10.1109/ACCESS.2020.2981745
4	WEI F, CHEN W, WU Y P, et al. Toward 5G wireless interface technology: enabling nonorthogonal multiple access in the sparse code domain [J]. IEEE vehicular technology magazine, 2018, 13(4): 18–27. DOI: 10.1109/MVT.2018.2867280
5	LIU L, YU W. Massive connectivity with massive MIMO: part I: device activity detection and channel estimation [J]. IEEE transactions on signal processing, 2018, 66(11): 2933–2946. DOI: 10.1109/TSP.2018.2818082
6	DING J, QU D M, FENG M J, et al. Dynamic preamble-resource partitioning for critical MTC in massive MIMO systems [J]. IEEE Internet of Things journal, 2021, 8(20): 15361–15371. DOI: 10.1109/JIOT.2021.3064061
7	RUSEK F, PERSSON D, LAU B K, et al. Scaling up MIMO: opportunities and challenges with very large arrays [J]. IEEE signal processing magazine, 2013, 30(1): 40–60. DOI: 10.1109/MSP.2011.2178495
8	SHEN J C, ZHANG J, CHEN K C, et al. High-dimensional CSI acquisition in massive MIMO: Sparsity-inspired approaches [J]. IEEE systems journal, 2017, 11(1): 32–40. DOI: 10.1109/JSYST.2015.2448661
9	LIU L, LARSSON E G, YU W, et al. Sparse signal processing for grant-free massive connectivity: a future paradigm for random access protocols in the Internet of Things [J]. IEEE signal processing magazine, 2018, 35(5): 88–99. DOI: 10.1109/MSP.2018.2844952
10	DAI J S, LIU A, LAU V K N. FDD massive MIMO channel estimation with arbitrary 2D-array geometry [J]. IEEE transactions on signal processing, 66(10): 2584–2599. DOI: 10.1109/tsp.2018.2807390
11	RAJORIYA A, RUKHSANA S, BUDHIRAJA R. Centralized and decentralized active user detection and channel estimation in mMTC [J]. IEEE transactions on communications, 70(3): 1759–1776. DOI: 10.1109/tcomm.2022.3141401
12	JALAL A, ARVINTE M, DARAS G, et al. Robust compressed sensing MRI with deep generative priors [C]//Proc. 35th International Conference on Neural Information Processing Systems. ACM, 2021: 14938–14954. DOI: 10.5555/3540261.3541406
13	SONG Y, SHEN L Y, XING L, et al. Solving inverse problems in medical imaging with score-based generative models [EB/OL]. (2021-11-15)[2024-08-30].
14	MENG X M, KABASHIMA Y. Diffusion model based posterior sampling for noisy linear inverse problems [EB/OL]. (2022-11-20)[2024-12-11].
15	CANDÈS E J, ROMBERG J K, TAO T. Stable signal recovery from incomplete and inaccurate measurements [J]. Communications on pure and applied mathematics, 2006, 59(8): 1207–1223. DOI: 10.1002/cpa.20124
16	WU Y, ROSCA M, LILLICRAP T. Deep compressed sensing [C]//36th International Conference on Machine Learning. PMLR, 2019: 6850–6860
17	AALI A, ARVINTE M, KUMAR S, et al. Solving inverse problems with score-based generative priors learned from noisy data [C]//Proc. 57th Asilomar Conference on Signals, Systems, and Computers. IEEE, 2023: 837–843. DOI: 10.1109/IEEECONF59524.2023.10477042
18	ALKHATEEB A, LEUS G, HEATH R W. Compressed sensing based multi-user millimeter wave systems: how many measurements are needed? [C]//Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2015: 2909–2913.. DOI: 10.1109/icassp.2015.7178503
19	REZENDE D J, MOHAMED S, REZENDE D J, et al. Variational inference with normalizing flows [C]//Proc. 32nd International Conference on International Conference on Machine Learning - Volume 37. ACM, 2015: 1530–1538. DOI: 10.5555/3045118.3045281
20	SONG Y, ERMON S. Generative modeling by estimating gradients of the data distribution [J]. 2019: 11918–11930. DOI: 10.5555/3454287.3455354
21	SONG Y, ERMON S, SONG Y, et al. Improved techniques for training score-based generative models [C]//Proc. 34th International Conference on Neural Information Processing Systems. ACM, 2020: 12438–12448. DOI: 10.5555/3495724.3496767
22	MENG X M, KABASHIMA Y. Quantized compressed sensing with score-based generative models [EB/OL]. (2022-11-02)[2024-12-11].
23	BORA A, JALAL A, PRICE E, et al. Compressed sensing using generative models [C]//Proc. 34th International Conference on Machine Learning. ACM, 2017: 537–546. DOI: 10.5555/3305381.3305437
24	HASAN M, HOSSAIN E, NIYATO D. Random access for machine-to-machine communication in LTE-advanced networks: issues and approaches [J]. IEEE communications magazine, 2013, 51(6): 86–93. DOI: 10.1109/MCOM.2013.6525600
25	PRATAS N K, THOMSEN H, STEFANOVIĆ Č, et al. Code-expanded random access for machine-type communications [C]//Proc. IEEE Globecom Workshops. IEEE, 2012: 1681–1686. DOI: 10.1109/GLOCOMW.2012.6477838
26	CHEN Z L, YU W. Massive device activity detection by approximate message passing [C]//Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2017: 3514–3518. DOI: 10.1109/ICASSP.2017.7952810
27	DONOHO D L, MALEKI A, MONTANARI A. Message-passing algorithms for compressed sensing [J]. Proceedings of the national academy of sciences of the United States of America, 2009, 106(45): 18914–18919. DOI: 10.1073/pnas.0909892106
28	BAYATI M, MONTANARI A. The dynamics of message passing on dense graphs, with applications to compressed sensing [J]. IEEE transactions on information theory, 2011, 57(2): 764–785. DOI: 10.1109/TIT.2010.2094817
29	HO J, JAIN A, ABBEEL P. Denoising diffusion probabilistic models [C]//34th International Conference on Neural Information Processing Systems. NIPS, 2020: 6840–6851
30	BATZOLIS G, STANCZUK J, SCHÖNLIEB C B, et al. Conditional image generation with score-based diffusion models [EB/OL]. (2021-11-26)[2024-12-11].
31	SOLTANI M, POURAHMADI V, MIRZAEI A, et al. Deep learning-based channel estimation [J]. IEEE communications letters, 2019, 23(4): 652–655. DOI: 10.1109/LCOMM.2019.2898944
32	HE H T, WEN C K, JIN S, et al. Deep learning-based channel estimation for beamspace mmWave massive MIMO systems [J]. IEEE wireless communications letters, 2018, 7(5): 852–855. DOI: 10.1109/LWC.2018.2832128
33	METZLER C A, MOUSAVI A, BARANIUK R G, et al. Learned D-amp [C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. ACM, 2017: 1770–1781. DOI: 10.5555/3294771.3294940
34	ARVINTE M, TAMIR J I. MIMO channel estimation using score-based generative models [J]. IEEE transactions on wireless communications, 2023, 22(6): 3698–3713. DOI: 10.1109/TWC.2022.3220784
35	LARSSON E G, EDFORS O, TUFVESSON F, et al. Massive MIMO for next generation wireless systems [J]. IEEE communications magazine, 52(2): 186–195. DOI: 10.1109/mcom.2014.6736761
36	BORGES D, MONTEZUMA P, DINIS R, et al. Massive MIMO techniques for 5G and beyond: opportunities and challenges [J]. Electronics, 2021, 10(14): 1667. DOI: 10.3390/electronics10141667
37	TARANTOLA A. Inverse problem theory and methods for model parameter estimation [M]. Philadelphia, USA: Society for Industrial and Applied Mathematics, 2005. DOI: 10.1137/1.9780898717921
38	BOTTARELLI M, EPIPHANIOU G, ISMAIL D K BEN, et al. Physical characteristics of wireless communication channels for secret key establishment: a survey of the research [J]. Computers & security, 2018, 78: 454–476. DOI: 10.1016/j.cose.2018.08.001
39	VINCENT P. A connection between score matching and denoising autoencoders [J]. Neural computation, 2011, 23(7): 1661–1674. DOI: 10.1162/neco_a_00142
40	SONG Y, DURKAN C, MURRAY I, et al. Maximum likelihood training of score-based diffusion models [EB/OL]. (2021-01-22)[2024-08-30].
41	ROBERTS G O, TWEEDIE R L. Exponential convergence of Langevin distributions and their discrete approximations [J]. Bernoulli, 1996, 2(4): 341. DOI: 10.2307/3318418
42	DANUFANE F H, DI RENZO M, DE ROSNY J, et al. On the path-loss of reconfigurable intelligent surfaces: an approach based on Green's theorem applied to vector fields [J]. IEEE transactions on communications, 2021, 69(8): 5573–5592. DOI: 10.1109/TCOMM.2021.3081452
43	DATTA T, KUMAR N A, CHOCKALINGAM A, et al. A novel Monte-Carlo-sampling-based receiver for large-scale uplink multiuser MIMO systems [J]. IEEE transactions on vehicular technology, 2013, 62(7): 3019–3038. DOI: 10.1109/TVT.2013.2260572
44	AOUDIA F AIT, HOYDIS J. End-to-end learning for OFDM: from neural receivers to pilotless communication [J]. IEEE transactions on wireless communications, 2022, 21(2): 1049–1063. DOI: 10.1109/TWC.2021.3101364

[1]	CHENG Jiaming, CHEN Wei, LI Lun, AI Bo. Efficient Spatio-Temporal Predictive Learning for Massive MIMO CSI Prediction [J]. ZTE Communications, 2025, 23(1): 3-10.
[2]	WANG Jilin, ZENG Xianlong, YANG Yonghui, PENG Lin, LI Lingxiang. Tensor Decomposition-Based Channel Estimation and Sensing for Millimeter Wave MIMO-OFDM V2I Systems [J]. ZTE Communications, 2024, 22(3): 56-68.
[3]	XIE Xinyu, WU Yongpeng, YUAN Zhifeng, MA Yihua. Massive Unsourced Random Access Under Carrier Frequency Offset [J]. ZTE Communications, 2023, 21(3): 45-53.
[4]	GUAN Xinrong, WU Qingqing. IRS‑Enabled Spectrum Sharing: Interference Modeling, Channel Estimation and Robust Passive Beamforming [J]. ZTE Communications, 2022, 20(1): 28-35.
[5]	SHAO Zhichao, YAN Wenjing, YUAN Xiaojun. Markovian Cascaded Channel Estimation for RIS Aided Massive MIMO Using 1‑Bit ADCs and Oversampling [J]. ZTE Communications, 2022, 20(1): 48-56.
[6]	JIAN Mengnan, ZHANG Nan, CHEN Yijian. RIS: Spatial‑Wideband Effect Analysis and Off‑Grid Channel Estimation [J]. ZTE Communications, 2022, 20(1): 57-62.
[7]	NAIKOTI Ashwitha, CHOCKALINGAM Ananthanarayanan. Signal Detection and Channel Estimation in OTFS [J]. ZTE Communications, 2021, 19(4): 16-33.
[8]	Julian AHRENS, Lia AHRENS, Hans D. SCHOTTEN. A Machine Learning Method for Prediction of Multipath Channels [J]. ZTE Communications, 2019, 17(4): 12-18.
[9]	Stefano Buzzi, Carmen D’Andrea. Massive MIMO 5G Cellular Networks: mm-Wave vs. μ-Wave Frequencies [J]. ZTE Communications, 2017, 15(S1): 41-49.
[10]	ZHANG Jianhua, WANG Chao, WU Zhongyuan, ZHANG Weite. A Survey of Massive MIMO Channel Measurements and Models [J]. ZTE Communications, 2017, 15(1): 14-22.
[11]	ZHANG Ping, CHEN Jianqiao, TANG Tian. An Overview of Non-Stationary Property for Massive MIMO Channel Modeling [J]. ZTE Communications, 2017, 15(1): 3-7.
[12]	Sohail Taheri, Mir Ghoraishi, XIAO Pei, CAO Aijun, GAO Yonghong. Evaluation of Preamble Based Channel Estimation for MIMO-FBMC Systems [J]. ZTE Communications, 2016, 14(4): 3-10.
[13]	Yifei Yuan, Xiaowu Zhao. 5G: Vision, Scenarios and Enabling Technologies [J]. ZTE Communications, 2015, 13(1): 3-10.
[14]	Fa-Long Luo. Signal Processing Techniques for 5G: An Overview [J]. ZTE Communications, 2015, 13(1): 20-27.

Device Activity Detection and Channel Estimation Using Score-Based Generative Models in Massive MIMO

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 6

References 44

Related Articles 14

Recommended Articles 0

Metrics