A Compressed Tensor-based Subspace Framework for 2D-DOA and Polarization Estimation
Liangliang Li, Xianpeng Wang, Xiang Lan · 2024
In this article, a compressed tensor-based subspace framework is developed to solve the parameter estimation problem in millimeter wave (mmWave) polarized massive multiple-input multiple-output (MIMO) architectures. In the proposed approach, the array measurement is first stacked into a high-dimensional third-order tensor to capture the tensor gain. Then, the high-dimensional third-order tensor is compressed into a low-dimensional one via a tensor compressive sampling (TCS) framework. Subsequently, the tensor-based signal subspace is achieved by imposing higher-order singular value decomposition (HOSVD) on the compressed tensor. Afterwards, the normalized polarization response vector is calculated exploiting the rotational invariance property. Lastly, two-dimensional (2D) direction-of-arrival (DOA) and polarization estimation are implemented with the help of vector cross-product and least squares (LS) techniques, respectively. By incorporating tensor gain with CS, the developed method computationally economical offers acceptable estimation accuracy. Meanwhile, it's insensitive to the sensor position. Numerical simulations demonstrate the advantages of the designed scheme.