[论文解读] On the Degrees of Freedom and Eigenfunctions of Line-of-Sight Holographic MIMO Communications

Juan Carlos Ruiz-Sicilia, Marco Di Renzo+3 more|arXiv (Cornell University)|Aug 15, 2023

Advanced Antenna and Metasurface Technologies被引用 5

一句话总结

本文提出了一种基于四次波前近似的可视线全息MIMO系统中有效自由度（eDoF）及最优通信波形的闭式表达式，克服了抛物线近似的局限性。该方法将eDoF表述为一个带限核的兰道特征值问题实例，从而实现对非傍轴部署中正交通信模式及最优基函数的解析计算。

ABSTRACT

We consider a line-of-sight communication link between two holographic surfaces (HoloSs), and provide a closed-form expression for the effective degrees of freedom (eDoF), i.e., the number of orthogonal communication modes that can be established between them. The proposed framework can be applied to network deployments beyond the widely studied paraxial setting. This is obtained by partitioning the largest HoloS into sub-HoloSs, and proving that the supports of the Fourier transforms of the kernels of the obtained integral operators are limited and are almost disjoint in the wavenumber domain, provided that the sub-HoloSs are sufficiently small. Using the proposed approach, it is proved that (i) the eDoF correspond to an instance of Landau's second eigenvalue problem; (ii) the eigenvalues polarize asymptotically to multiple values; and (iii) the eDoF depend explicitly on the approximation accuracy according to Kolmogorov's n-width criterion. This result generalizes the analysis in the paraxial setting, in which it is known that the eigenvalues polarize asymptotically to two values. In addition, it is proved that the typical method of analysis utilized in the paraxial setting, which is based on a parabolic approximation of the wavefront in a local coordinates system, is equivalent to a quartic approximation of the wavefront in a general coordinates system. This facilitates the derivation of an explicit formula for the eDoF in terms of key system parameters, including the relative offset between the center-points of the HoloSs, and their relative rotation and tilt. We specialize the framework to canonical network deployments, and provide analytical expressions for the optimal, according to Kolmogorov's n-width criterion, basis functions (communication waveforms) for data encoding and decoding.

研究动机与目标

• 克服在非傍轴部署中使用抛物线近似估算全息MIMO自由度（DoF）时的不准确性。
• 推导可视线全息MIMO系统中有效自由度（eDoF）的闭式表达式。
• 基于柯尔莫哥洛夫N-宽度准则，解析确定用于数据编码与解码的最优基函数（通信波形）。
• 将框架推广至非傍轴区域，适用于任意网络部署。
• 通过eDoF与波形精度的数值分析验证解析结果。

提出的方法

• 采用电磁波前的四次近似来建模信道核，使解析处理超越抛物线近似成为可能。
• 将eDoF计算重新表述为源自四次波前模型的带限核的兰道特征值问题。
• 在希尔伯特空间中应用紧算子与自伴算子理论，推导出对应于最优通信波形的特征函数。
• 将全息面划分为子全息面（sub-HoloSs），并在波数域中使用傅里叶变换计算核支撑集的勒贝格测度。
• 通过子全息面及其傅里叶变换核的勒贝格测度的乘积，推导出eDoF的上界。
• 通过变量分离求解所得特征值问题，将其简化为可通过掠线球面波函数（PSWFs）求解的一维sinc型积分方程。

实验结果

研究问题

• RQ1 在非傍轴、四次波前近似下，可视线全息MIMO链路中的有效自由度（eDoF）数量是多少？
• RQ2 全息MIMO中通信波形的最优基函数与信道核的特征函数有何关系？
• RQ3 是否可以在不依赖数值特征值求解器的情况下，以闭式表达计算eDoF与波形？
• RQ4 四次波前近似在非傍轴部署中相比抛物线近似如何提升精度？
• RQ5 全息面的空间分割与所得eDoF及波形正交性之间存在何种关系？

主要发现

• 有效自由度（eDoF）的数量以闭式表达，即四次波前近似所得带限核的兰道特征值问题的解。
• 发射与接收的最优基函数被证明为一维积分算子的可分离特征函数，可通过掠线球面波函数（PSWFs）求解。
• eDoF的上界为(2π)⁻²乘以傅里叶变换后子全息面的勒贝格测度之和，考虑了波数域中支撑集的重叠。
• 该框架实现了在非傍轴场景中eDoF与波形的解析计算，而此时抛物线近似失效。
• 数值验证确认了eDoF闭式表达式的准确性，以及所推导波形的正交性。
• 所推导的波形最小化了柯尔莫哥洛夫N-宽度近似误差，确保在给定误差容限内的最优数据表示。