基于离格Turbo变分贝叶斯宽带信号到达角估计  

Wideband DOA Estimation for Off-grid Sources Using the Turbo-VBI Algorithm

作　　者：瞿扬 杨竞舟 李大鹏[1] QU Yang;YANG Jingzhou;LI Dapeng(School of Communication and Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210003,China;Institute of Microelectronics of the Chinese Academy of Sciences,Beijing 100029,China)

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]中国科学院微电子研究所,北京100029

出　　处：《无线电工程》2025年第2期221-229,共9页Radio Engineering

基　　金：国家自然科学基金(62371245)。

摘　　要：与传统均匀线阵相比,嵌套阵列在相同数量物理天线的条件下,能够实现更大的阵列孔径和更丰富的自由度。在面向嵌套阵列的宽带信号应用中,针对传统的稀疏贝叶斯(Sparse Bayesian Learning, SBL)算法在解决来波方向(Direction of Arrival, DOA)估计中存在实际的DOA角度无法完全落在离散化的网格上的问题,提出了一种离格(Off-Grid, OG)Turbo变分贝叶斯(Variational Bayesian Inference, VBI)算法。通过带通滤波器组或离散傅里叶变换(Discrete Fourier Transform, DFT)将宽带信号分解为多个窄带信号,每个子带的角度域稀疏性可以建模为复高斯分布,在VBI框架下实现OG的高分辨率DOA估计。仿真结果表明,与其他压缩感知算法相比,该算法具有高精度、易收敛的特性;在少快拍数或者单快拍数下能够精确定位DOA,具有高鲁棒性。Compared to traditional uniform linear arrays,nested arrays can achieve a larger array aperture and offer richer degrees of freedom with the same number of physical antennas.In wideband signal applications for nested arrays,to address the problem in traditional Sparse Bayesian Learning(SBL)where the actual Direction of Arrival(DOA)angles may not exactly fall on the discretized grid,an Off-Grid(OG)Turbo Variational Bayesian Interference(VBI)algorithm is proposed.By using band-pass filter banks or Discrete Fourier Transform(DFT),the wideband signal is decomposed into multiple narrowband signals.The angular domain sparsity of each subband can be modeled as a complex Gaussian distribution,allowing OG high-resolution DOA estimation within the variational Bayesian framework.Simulations demonstrate that,compared to other compressed sensing algorithms,the proposed method has the advantages of high accuracy and fast convergence.Moreover,the algorithm can precisely estimate DOA even with a small number of snapshots or a single snapshot,indicating its high robustness.

关 键 词：稀疏贝叶斯算法 嵌套阵列 来波方向估计 压缩感知 

分 类 号：TN911.23[电子电信—通信与信息系统]