面向船载低轨卫星通信终端的闭环反馈式自适应对星算法

陈浩天; 马紫娴; 谢信洪; 李娜雨; 李保珠; 宋春毅; 徐志伟

doi:10.11999/JEIT251324

面向船载低轨卫星通信终端的闭环反馈式自适应对星算法

doi: 10.11999/JEIT251324 cstr: 32379.14.JEIT251324

陈浩天^{1, 3},
马紫娴^{1, 3},
谢信洪²,
李娜雨^{2, 1},
李保珠²,
宋春毅^{2, 1, 3, ,},
徐志伟^{2, 1, 3}

1.
浙江大学海洋精准感知技术全国重点实验室，& 海洋学院，浙江大学舟山 316021
2.
东海实验室舟山 316021
3.
海洋感知技术与装备教育部工程研究中心舟山 316021

基金项目: 浙江省重点研发计划项目(项目编号：2024SSYS0095)

详细信息

作者简介:
陈浩天：男，博士生，研究方向为无线通信领域的射频及毫米波相控阵系统，以及波束调控算法

马紫娴：女，博士生，研究方向为无线通信领域的射频及毫米波相控阵系统

谢信洪：男，硕士，研究方向为无线通信领域的射频及毫米波相控阵系统

李娜雨：男，博士，副研究员，研究方向为集成电路与系统

李保珠：男，博士，副研究员，研究方向为车联网、卫星互联网、机器学习、资源管理

宋春毅：男，博士，教授，研究方向为低轨卫星宽带通信、智能无障感知(无人船/艇、无人驾驶)、人工智能标准

徐志伟：男，博士，教授，研究方向为物联网与通信应用的集成电路与系统

通讯作者:
宋春毅　cysong@zju.edu.cn

中图分类号: TN927.2
计量
- 文章访问数: 7
- HTML全文浏览量: 3
- PDF下载量: 0
- 被引次数: 0
出版历程
- 修回日期: 2026-03-03
- 录用日期: 2026-03-03
- 网络出版日期: 2026-03-14

A Closed-loop Feedback Adaptive Beam Alignment Algorithm for Shipborne LEO Satellite Communication Terminals

CHEN Haotian^{1, 3},
MA Zixian^{1, 3},
XIE Xinhong²,
LI Nayu^{2, 1},
LI Baozhu²,
SONG Chunyi^{2, 1, 3
, ,},
XU Zhiwei^{2, 1, 3}

1.
The State Key Laboratory of Ocean Sensing and Ocean College, Zhejiang University, Zhoushan 316021, China
2.
Donghai Laboratory, Zhoushan 316021, China
3.
The Engineering Research Center of Oceanic Sensing Technology and Equipment, Ministry of Education, Zhoushan 316021, China

Funds: The Key R&D Program of Zhejiang Province (Grant 2024SSYS0095)

摘要

摘要: 海上风浪扰动导致船舶与低轨(LEO)卫星间产生非线性相对运动，其复杂动态特性加剧了船载相控阵终端与星载相控阵终端间实现波束对准的技术难度，成为制约海洋低轨卫星通信稳定性与可靠性的关键瓶颈。针对船舶与低轨卫星构成的“动动通”通信场景，本文提出了一种闭环反馈式自适应对星算法：采用块稀疏贝叶斯学习(block-SBL)算法实现基于粗糙网格的快速目标捕获，再通过自适应牛顿迭代法提升离网估计精度；在此基础上，将离网估计结果输入无迹卡尔曼滤波器(UKF)，预测并补偿船体在算法处理时延内新产生的姿态偏差，最终形成捕获与追踪协同的闭环反馈式自适应对星算法。基于真实海况下28,000-DWT级散货船实测运动姿态数据的仿真结果表明，在高海况条件下，相较于现有的离网波达方向(DOA)估计算法，所提算法不仅能有效校准估计结果，其计算复杂度亦显著降低。
- 低轨卫星通信 /
- 到达角估计 /
- 多目标追踪 /
- 压缩感知 /
- 稀疏贝叶斯学习
Abstract: Objective The 6G-based satellite communication (SATCOM) network has emerged as the primary solution for achieving ubiquitous communication and oceanic communications. Compared with traditional geostationary earth orbit (GEO) satellites, the latest generation of low earth orbit (LEO) satellites offer higher throughput, lower end-to-end latency, and reduced deployment costs. Thus, phased arrays have been widely adopted in LEO SATCOM due to their beam agility. However, maritime wind-wave disturbances induce nonlinear relative motion between shipborne terminals and LEO satellites, posing severe challenges for high-precision satellite acquisition and tracking. To solve this problem, LEO SATCOM systems require a novel beam alignment algorithm that first obtains the target’s instantaneous state and motion characteristics through a target acquisition algorithm, then utilizes multi-target tracking method to predict satellite trajectories based on target states, compensating for estimation errors caused by severe coupled motions. Methods The proposed closed-loop feedback adaptive beam alignment algorithm consists of two tightly coupled components: target acquisition and target state updating. In the target acquisition stage, a rank reduction estimator (RARE) is first employed to decompose the array factor matrix and transform the original two-dimensional direction of arrival (DOA) estimation problem into two sequential one-dimensional estimations, which significantly reduces the computational complexity of each sparse Bayesian learning (SBL) iteration. Based on the coarse grid obtained from RARE, an adaptive Newton sparse Bayesian learning (ANSBL) method is developed. ANSBL utilizes block-sparse Bayesian learning to achieve initial target acquisition on the coarse grid and subsequently performs two-stage Newton refinement to mitigate off-grid mismatch, thereby providing high-accuracy DOA estimation in both $ \theta $ and $ \varphi $ and effectively improving angular observation precision. In the target state updating stage, an unscented Kalman filter (UKF)-based ternary joint prediction mechanism is introduced. The UKF simultaneously predicts the target’s motion state, the signal variance, and the noise variance for next target acquisition process. These predicted probability distributions are then used to update the initial grid and hyperparameters of the subsequent SBL acquisition stage, providing more consistent and comprehensive initial values. Through this closed-loop interaction, target acquisition and state tracking are deeply integrated, which substantially reduces the number of SBL iterations required for convergence. This improvement is particularly significant under high sea-state conditions where reducing beam alignment time is essential. Results and Discussions The proposed closed-loop feedback adaptive beam alignment algorithm first uses on-grid DOA estimation to reduce array factor correlation and improve efficiency during target acquisition, then utilizes Newton iteration for higher off-grid accuracy (Fig. 3). Subsequently, the proposed method was validated using real-world ship attitude data collected from a 28,000-DWT bulk carrier under actual sea conditions (Fig. 4). The UKF refines DOA results via state updating, with its prediction of signal position, signal variance, and noise variance providing accurate hyperparameter initial values, thus reducing iterations and converging faster than other algorithms (Figs. 5). Under low sea-state conditions, the proposed method not only achieves satellite alignment in less than 0.2 seconds, but also reduces the satellite position estimation error from ±1° to ±0.5° (Fig. 6(a)). Furthermore, under high sea-state conditions, the UKF effectively predicts satellite positions; more importantly, it cuts the satellite position estimation error from ±2.5° to ±0.65°, fully verifying the method’s robust tracking accuracy and error mitigation capability in harsh marine environments (Fig. 6(b)). Conclusions To address the performance requirements of beam alignment algorithms for LEO communication satellites, this paper proposes a closed-loop feedback adaptive beam alignment algorithm. The algorithm first employs a block-based SBL algorithm to obtain grid-based DOA estimation results, and achieves super-resolution direction estimation under off-grid conditions via adaptive Newton iteration. Through the UKF, the estimation results are dynamically calibrated in real time. The UKF further predicts the target’s motion state, the signal variance, and the noise variance for the next target acquisition process, thereby enhancing tracking continuity and alignment accuracy. Numerical simulations demonstrate that the proposed algorithm outperforms traditional beam alignment methods in both numerical accuracy and robustness, effectively mitigating severe shaking interference from terminals under complex sea conditions.
- LEO satellite communication (SATCOM) /
- Direction of arrival (DOA) /
- Multi-target tracking /
- Compressed sensing /
- Sparse Bayesian learning

HTML全文

图 1 Walker星座网络示意图

下载: 全尺寸图片幻灯片

图 2 相控阵通信终端六自由度运动模型示意图

下载: 全尺寸图片幻灯片

图 3 DOA估计性能对比图

下载: 全尺寸图片幻灯片

图 4 28,000-DWT级散货船横摇角、纵摇角和艏摇角随时间变化图

下载: 全尺寸图片幻灯片

图 5 基于28,000-DWT级散货船实测姿态数据的算法追踪时间对比图

下载: 全尺寸图片幻灯片

图 6 基于28,000-DWT级散货船实测姿态数据的卫星追踪性能对比图

下载: 全尺寸图片幻灯片

表 1 SU-ANSBL流程表

SU-ANSBL
1.初始化：
迭代参数：追踪时刻$ t=0 $；追踪总次数：$ {T}_{\text{track}} $；追踪间隔$ {\Delta }_{t} $
实施ANSBL，获得初始状态矩阵：$ {\boldsymbol{E}}_{p,0} $、$ {\boldsymbol{E}}_{\boldsymbol{\delta },0} $、$ {\boldsymbol{E}}_{\sigma ,0} $和初始方差矩阵：$ {\boldsymbol{P}}_{p,0} $、$ {\boldsymbol{P}}_{\boldsymbol{\delta },0} $、$ {\boldsymbol{P}}_{\sigma ,0} $
2. SU-ANSBL：
当$ t\leq {T}_{\text{track}} $时：
更新各均值矩阵和方差矩阵的先验预测值：$ {{\overline{\boldsymbol{\chi }}}}_{p,t\|t-1} $、$ {{\overline{\boldsymbol{\chi }}}}_{\boldsymbol{\delta },t\|t-1} $、$ {{\overline{\boldsymbol{\chi }}}}_{\sigma ,t\|t-1} $、$ {\boldsymbol{P}}_{p,t\|t-1} $、$ {\boldsymbol{P}}_{\boldsymbol{\delta },t\|t-1} $、$ {\boldsymbol{P}}_{\sigma ,t\|t-1} $
更新字典矩阵$ {\boldsymbol{\boldsymbol{\varPhi }}}_{x} $和$ {{\boldsymbol{\varPhi }}}_{y} $，信号方差向量$ {\boldsymbol{\delta }}_{x} $和$ {\boldsymbol{\delta }}_{y} $，噪声方差$ {\sigma }^{2} $
实施ANSBL，获得DOA估计结果，计算观测矩阵：$ {{\tilde{\boldsymbol{Z}}}}_{p,\text{t}} $、$ {{\tilde{\boldsymbol{Z}}}}_{\boldsymbol{\delta },0} $、$ {{\tilde{\boldsymbol{Z}}}}_{\sigma ,0} $
更新各均值矩阵和方差矩阵的后验预测值：$ {{\overline{\boldsymbol{\chi }}}}_{p,t} $、$ {{\overline{\boldsymbol{\chi }}}}_{\boldsymbol{\delta },t} $、$ {{\overline{\boldsymbol{\chi }}}}_{\sigma ,t} $、$ {\boldsymbol{P}}_{p,t} $、$ {\boldsymbol{P}}_{\boldsymbol{\delta },t} $、$ {\boldsymbol{P}}_{\sigma ,t} $
根据$ {{\overline{\boldsymbol{\chi }}}}_{p,t} $计算信号源位置的最终估计值并输出：$ \left({\tilde{\theta }}_{k},{\tilde{\varphi }}_{k}\right) $
$ t\leftarrow t+{\Delta }_{t} $

下载: 导出CSV

表 2 28,000-DWT 级散货船，Walker星座参数表

名称	基本参数	数值
28,000-DWT 级散货船	垂直间长	160.4 m
	船宽	27.2 m
	平均吃水	8.16 m
	常用航速	12节
Walker星座	$ N/P/F $	160/8/0
	卫星高度	1200 km
	轨道倾角	$ {60}^{\circ } $

下载: 导出CSV

参考文献(25)

[1]	冯伊凡, 吴畏虹, 孙罡, 等. 天地一体化边缘计算网络服务迁移算法研究[J]. 电子与信息学报, 2026, 48(2): 499–511. doi: 10.11999/JEIT250835. FENG Yifan, WU Weihong, SUN Gang, et al. Service migration algorithm for satellite-terrestrial edge computing networks[J]. Journal of Electronics & Information Technology, 2026, 48(2): 499–511. doi: 10.11999/JEIT250835.
[2]	GUO Ningxuan, LIU Liang, and ZHONG Xiaoqing. Task-aware distributed inter-layer topology optimization method in resource-limited LEO-LEO satellite networks[J]. IEEE Transactions on Wireless Communications, 2024, 23(4): 3572–3585. doi: 10.1109/TWC.2023.3309379.
[3]	LIU Ying, ZHANG Wenbo, JIA Yongtao, et al. Low RCS antenna array with reconfigurable scattering patterns based on digital antenna units[J]. IEEE Transactions on Antennas and Propagation, 2021, 69(1): 572–577. doi: 10.1109/TAP.2020.3004993.
[4]	梁广, 龚文斌, 余金培. 基于子阵列的低轨星载多波束相控阵天线的设计与实现[J]. 电子与信息学报, 2010, 32(6): 1435–1440. doi: 10.3724/SP.J.1146.2009.00695. LIANG Guang, GONG Wenbin, and YU Jinpei. The design and implementation of sub-arrayed phased array antenna for LEO satellite[J]. Journal of Electronics & Information Technology, 2010, 32(6): 1435–1440. doi: 10.3724/SP.J.1146.2009.00695.
[5]	WÜNSCHE R, KRONDORF M, and KNOPP A. Investigations of channel capacity loss in LEO satellite systems using phased array beamforming antennas[J]. IEEE Transactions on Aerospace and Electronic Systems, 2024, 60(4): 5373–5394. doi: 10.1109/TAES.2024.3392181.
[6]	LI Min, GAO Huiyan, LI Nayu, et al. A 17.7–19.2-GHz receiver front end with an adaptive analog temperature- compensation scheme[J]. IEEE Transactions on Microwave Theory and Techniques, 2023, 71(3): 1068–1082. doi: 10.1109/TMTT.2022.3215558.
[7]	LI Nayu, MA Zixian, XIE Xinhong, et al. A Ka-band 64-element four-beam phased array receiver with inter-beam interference cancellation[J]. IEEE Transactions on Microwave Theory and Techniques, 2025, 73(1): 221–233. doi: 10.1109/TMTT.2024.3418879.
[8]	HONG Ru, QI Xiaokang, XIE Xinhong, et al. A dual-beam 256-element K-band Active phased array with reconfigurable polarization[J]. IEEE Antennas and Wireless Propagation Letters, 2026. doi: 10.1109/LAWP.2026.3661753. (查阅网上资料,未找到卷期页码信息,请确认补充).
[9]	MA Zixian, XIE Xinhong, LI Nayu, et al. Multibeam-assisted cross-polarization suppression with Ka-band 64-element four-beam reconfigurable phased array implementation[J]. IEEE Transactions on Microwave Theory and Techniques, 2025, 73(12): 10814–10826. doi: 10.1109/TMTT.2025.3612979.
[10]	XIE Xinhong, CHEN Haotian, MA Zixian, et al. A power-only fast calibration method for phased array using convolution neural network[J]. IEEE Antennas and Wireless Propagation Letters, 2024, 23(11): 3972–3976. doi: 10.1109/LAWP.2024.3452268.
[11]	CHEN Haotian, XIE Xinhong, MA Zixian, et al. A novel fast far-field phased array calibration method utilizing deep residual neural networks[J]. IEEE Transactions on Antennas and Propagation, 2025, 73(4): 2217–2231. doi: 10.1109/TAP.2025.3547915.
[12]	TAIRA S, TANAKA M, and OHMORI S. High gain airborne antenna for satellite communications[J]. IEEE Transactions on Aerospace and Electronic Systems, 1991, 27(2): 354–360. doi: 10.1109/7.78309.
[13]	CHEN Qin, XU Yuying, SONG Chunyi, et al. Adaptive tracking for beam alignment between ship-borne digital phased-array antenna and LEO satellite[J]. Journal of Communications and Information Networks, 2019, 4(3): 60–70. doi: 10.23919/JCIN.2019.8917886.
[14]	李卓. 面向复杂信号环境GNSS载波相位质量的深组合技术研究[D]. [硕士论文], 武汉大学, 2019. LI Zhuo. Research on deep integration technology for GNSS carrier phase quality in complex signal environment[D]. [Master dissertation], Wuhan University, 2019.
[15]	何红丽. 运动平台轨迹计算方法研究[J]. 国际航空航天科学, 2015, 3(2): 13–18. doi: 10.12677/JAST.2015.32002. HE Hongli. Study on trajectory algorithm of motion platform[J]. Journal of Aerospace Science and Technology, 2015, 3(2): 13–18. doi: 10.12677/JAST.2015.32002.
[16]	YUEN N and FRIEDLANDER B. Asymptotic performance analysis of ESPRIT, higher order ESPRIT, and virtual ESPRIT algorithms[J]. IEEE Transactions on Signal Processing, 1996, 44(10): 2537–2550. doi: 10.1109/78.539037.
[17]	ZHANG Chen, ZHAO Minjian, and CAI Yunlong. Joint 2-D DOA estimation and mutual coupling calibration for uniform rectangular arrays[C]. 2015 International Conference on Wireless Communications & Signal Processing (WCSP), Nanjing, China, 2015: 1–5. doi: 10.1109/WCSP.2015.7340991.
[18]	GANGULY S, GHOSH J, MUKHOPADHYAY M, et al. Multi-time snapshot based off-grid DOA estimation of sparse array antennas using MFOCUSS algorithm[C]. 2020 URSI Regional Conference on Radio Science (URSI-RCRS), Varanasi, India, 2020: 1–4. doi: 10.23919/URSIRCRS49211.2020.9113604.
[19]	CHOO Y and YANG H. Broadband off-grid DOA estimation using block sparse Bayesian learning for nonuniform noise variance[J]. IEEE Journal of Oceanic Engineering, 2022, 47(4): 1024–1040. doi: 10.1109/JOE.2022.3151949.
[20]	YANG Yang, CHU Zhigang, and PING Guoli. Two-dimensional multiple-snapshot grid-free compressive beamforming[J]. Mechanical Systems and Signal Processing, 2019, 124: 524–540. doi: 10.1016/j.ymssp.2019.02.011.
[21]	DAI Jisheng, BAO Xu, XU Weichao, et al. Root sparse Bayesian learning for off-grid DOA estimation[J]. IEEE Signal Processing Letters, 2017, 24(1): 46–50. doi: 10.1109/LSP.2016.2636319.
[22]	LIANG Guolong, LI Chenmu, QIU Longhao, et al. State-updating-based DOA estimation using sparse Bayesian learning[J]. Applied Acoustics, 2022, 192: 108719. doi: 10.1016/j.apacoust.2022.108719.
[23]	DING Yanwu, MINKLER C, and ZHANG Y D, et al. Single-satellite EMI geolocation via flexibly constrained UKF exploiting Doppler acceleration[J]. IEEE Signal Processing Letters, 2025, 32: 11–15. doi: 10.1109/LSP.2024.3487778.
[24]	宋佳蓁, 师卓越, 张晓平, 等. 基于快速无迹卡尔曼滤波的雷达高速目标追踪技术[J]. 电子与信息学报, 2025, 47(8): 2703–2713. doi: 10.11999/JEIT25001. SONG Jiazhen, SHI Zhuoyue, ZHANG Xiaoping, et al. Radar high-speed target tracking via quick unscented Kalman filter[J]. Journal of Electronics & Information Technology, 2025, 47(8): 2703–2713. doi: 10.11999/JEIT25001.
[25]	JING Qianfeng, SASA K, CHEN Chen, et al. Analysis of ship maneuvering difficulties under severe weather based on onboard measurements and realistic simulation of ocean environment[J]. Ocean Engineering, 2021, 221: 108524. doi: 10.1016/j.oceaneng.2020.108524.