JZUS - Journal of Zhejiang University SCIENCE

ENGINEERING Information Technology & Electronic Engineering

Accepted manuscript available online (unedited version)

Deep unfolding based channel estimation for wideband terahertz near-field massive MIMO systems

Author(s): Jiabao GAO, Xiaoming CHEN, Geoffrey Ye LI
Affiliation(s): College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou 310027, China; more
Corresponding email(s): gao_jiabao@zju.edu.cn, chen_xiaoming@zju.edu.cn, Geoffrey.Li@imperial.ac.uk
Key Words: Terahertz; Massive MIMO; Channel estimation; Deep learning

Share this article to： More <<< Previous Paper \|

Jiabao GAO, Xiaoming CHEN, Geoffrey Ye LI. Deep unfolding based channel estimation for wideband terahertz near-field massive MIMO systems[J]. Frontiers of Information Technology & Electronic Engineering,in press.https://doi.org/10.1631/FITEE.2300760

@article{title="Deep unfolding based channel estimation for wideband terahertz near-field massive MIMO systems",
author="Jiabao GAO, Xiaoming CHEN, Geoffrey Ye LI",
journal="Frontiers of Information Technology & Electronic Engineering",
year="in press",
publisher="Zhejiang University Press & Springer",
doi="https://doi.org/10.1631/FITEE.2300760"
}

%0 Journal Article
%T Deep unfolding based channel estimation for wideband terahertz near-field massive MIMO systems
%A Jiabao GAO
%A Xiaoming CHEN
%A Geoffrey Ye LI
%J Frontiers of Information Technology & Electronic Engineering
%P 1162-1172
%@ 2095-9184
%D in press
%I Zhejiang University Press & Springer
doi="https://doi.org/10.1631/FITEE.2300760"

TY - JOUR
T1 - Deep unfolding based channel estimation for wideband terahertz near-field massive MIMO systems
A1 - Jiabao GAO
A1 - Xiaoming CHEN
A1 - Geoffrey Ye LI
J0 - Frontiers of Information Technology & Electronic Engineering
SP - 1162
EP - 1172
%@ 2095-9184
Y1 - in press
PB - Zhejiang University Press & Springer
ER -
doi="https://doi.org/10.1631/FITEE.2300760"

Abstract
Chinese Summary
Academic Network
Reviewer Comment

Abstract: The combination of terahertz and massive multiple-input multiple-output (MIMO) is promising for meeting the increasing data rate demand of future wireless communication systems thanks to the significant bandwidth and spatial degrees of freedom. However, unique channel features, such as the near-field beam split effect, make channel estimation particularly challenging in terahertz massive MIMO systems. On one hand, adopting the conventional angular domain transformation dictionary designed for low-frequency far-field channels will result in degraded channel sparsity and destroyed sparsity structure in the transformed domain. On the other hand, most existing compressive sensing based channel estimation algorithms cannot achieve high performance and low complexity simultaneously. To alleviate these issues, in this study, we first adopt frequency-dependent near-field dictionaries to maintain good channel sparsity and sparsity structure in the transformed domain under the near-field beam split effect. Then, a deep unfolding based wideband terahertz massive MIMO channel estimation algorithm is proposed. In each iteration of the approximate message passing-sparse Bayesian learning algorithm, the optimal update rule is learned by a deep neural network (DNN), whose architecture is customized to effectively exploit the inherent channel patterns. Furthermore, a mixed training method based on novel designs of the DNN architecture and the loss function is developed to effectively train data from different system configurations. Simulation results validate the superiority of the proposed algorithm in terms of performance, complexity, and robustness.

基于深度展开的宽带太赫兹近场大规模天线信道估计

高佳宝¹，陈晓明¹，李烨²
¹浙江大学信息与电子工程学院，中国杭州市，310027
²伦敦帝国理工学院电子电气工程系，英国伦敦市，SW7 2BU
摘要：得益于巨大的带宽和空间自由度，太赫兹与大规模天线的结合有望满足未来无线通信系统不断增长的数据传输速率需求。然而，近场波束分裂效应等独特的信道特性使得太赫兹大规模天线系统的信道估计极具挑战性。一方面，采用针对低频远场信道设计的角度域变换字典会导致变换域信道稀疏度的下降并破坏其稀疏结构。另一方面，大多数现有基于压缩感知的信道估计算法无法同时取得高性能和低复杂度。为缓解这些问题，本文首先采用频率相关的近场字典以在近场波束分裂效应下维持良好的变换域信道稀疏度和稀疏结构。然后，提出一种基于深度展开的宽带太赫兹大规模天线信道估计算法。在近似消息传递-稀疏贝叶斯学习算法的每轮迭代中，通过一个深度神经网络学习最优更新规则，精心设计网络结构以有效利用内在信道规律。此外，开发了一种基于网络结构和损失函数设计的混合训练方法，以有效训练来自不同系统配置的数据。仿真结果验证了所提算法在性能、复杂度和鲁棒性上的优越性。

关键词组：太赫兹；大规模天线；信道估计；深度学习

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Reference

[1]Chen YH, Yan LF, Han C, 2021. Hybrid spherical- and planar-wave modeling and DCNN-powered estimation of terahertz ultra-massive MIMO channels. IEEE Trans Commun, 69(10):7063-7076.

[2]Cui MY, Dai LL, 2022. Channel estimation for extremely large-scale MIMO: far-field or near-field? IEEE Trans Commun, 70(4):2663-2677.

[3]Cui MY, Dai LL, 2023. Near-field wideband channel estimation for extremely large-scale MIMO. Sci China Inform Sci, 66(7):172303.

[4]Cui MY, Dai LL, Wang ZC, et al., 2023a. Near-field rainbow: wideband beam training for XL-MIMO? IEEE Trans Wirel Commun, 22(6):3899-3912.

[5]Cui MY, Tan JB, Dai LL, 2023b. Wideband hybrid precoding for THz massive MIMO with angular spread. Sci Sin Inform, 53(4):772-786.

[6]Elbir AM, Shi W, Papazafeiropoulos AK, et al., 2023. Near-field terahertz communications: model-based and model-free channel estimation. IEEE Access, 11:36409-36420.

[7]Gao JB, Hu M, Zhong CJ, et al., 2022. An attention-aided deep learning framework for massive MIMO channel estimation. IEEE Trans Wirel Commun, 21(3):1823-1835.

[8]Gao JB, Zhong CJ, Li GY, 2023a. AMP-SBL unfolding for wideband mmWave massive MIMO channel estimation. IEEE Int Conf on Communications Workshops, p.60-65.

[9]Gao JB, Zhong CJ, Li GY, et al., 2023b. Deep learning-based channel estimation for wideband hybrid mmWave massive MIMO. IEEE Trans Commun, 71(6):3679-3693.

[10]Hu XL, Liu CX, Peng MG, et al., 2023. IRS-based integrated location sensing and communication for mmWave SIMO systems. IEEE Trans Wirel Commun, 22(6):4132-4145.

[11]Lu Y, Dai LL, 2023. Near-field channel estimation in mixed LoS/NLoS environments for extremely large-scale MIMO systems. IEEE Trans Commun, 71(6):3694-3707.

[12]Luo M, Guo QH, Jin M, et al., 2021. Unitary approximate message passing for sparse Bayesian learning. IEEE Trans Signal Process, 69:6023-6039.

[13]Nayir H, Karakoca E, Görçin A, et al., 2022. Hybrid-field channel estimation for massive MIMO systems based on OMP cascaded convolutional autoencoder. Proc IEEE 96^th Vehicular Technology Conf, p.1-6.

[14]Qin ZJ, Ye H, Li GY, et al., 2019. Deep learning in physical layer communications. IEEE Wirel Commun, 26(2):93-99.

[15]Srivastava S, Mishra A, Rajoriya A, et al., 2019. Quasi-static and time-selective channel estimation for block-sparse millimeter wave hybrid MIMO systems: sparse Bayesian learning (SBL) based approaches. IEEE Trans Signal Process, 67(5):1251-1266.

[16]Wan ZW, Gao Z, Gao FF, et al., 2021. Terahertz massive MIMO with holographic reconfigurable intelligent surfaces. IEEE Trans Commun, 69(7):4732-4750.

[17]Wei XH, Dai LL, 2022. Channel estimation for extremely large-scale massive MIMO: far-field, near-field, or hybrid-field? IEEE Commun Lett, 26(1):177-181.

[18]Yu WT, Shen YF, He HT, et al., 2022. Hybrid far- and near-field channel estimation for THz ultra-massive MIMO via fixed point networks. IEEE Global Communications Conf, p.5384-5389.

[19]Zhang XY, Wang ZN, Zhang HY, et al., 2023. Near-field channel estimation for extremely large-scale array communications: a model-based deep learning approach. IEEE Commun Lett, 27(4):1155-1159.

[20]Zhu YF, Guo HY, Lau VKN, 2021. Bayesian channel estimation in multi-user massive MIMO with extremely large antenna array. IEEE Trans Signal Process, 69:5463-5478.

Open peer comments: Debate/Discuss/Question/Opinion

<1>

- Go to

基于深度展开的宽带太赫兹近场大规模天线信道估计

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Reference