Walsh Meets OAM in Holographic MIMO

Charles Vanwynsberghe, Jiguang He, Chongwen Huang, Mérouane Debbah

TL;DR

This paper introduces a novel communication scheme for holographic MIMO systems using polar Walsh functions, offering a hardware-friendly alternative to existing OAM-based methods. While showing a slight performance dip compared to OAM, it promises significant reductions in implementation complexity for future wireless networks.

What Problem Does This Paper Address?

Next-generation wireless networks are exploring Holographic MIMO, which utilizes large intelligent surfaces (LISs) with densely packed antennas to unlock massive spatial degrees of freedom and boost spectrum efficiency, especially in near-field communication scenarios. A key challenge in these systems is acquiring and processing the channel state information (CSI), which becomes computationally intensive and hardware-demanding with conventional methods like Singular Value Decomposition (SVD). Existing OAM-based precoding schemes offer a promising alternative by requiring less CSI and simpler hardware, but this paper aims to further simplify the hardware requirements by proposing a new precoding design based on polar Walsh functions.

Methods

The authors propose a new precoding design using polar Walsh functions for holographic MIMO systems with disk-shaped LISs. These functions are characterized by being separable in radial and angular coordinates, orthogonal, and having constant amplitude with only two phase values (binary phases). This binary phase characteristic is key to reducing hardware complexity compared to OAM modes, which use continuous phases. The paper analyzes the performance of this new scheme by comparing its channel capacity against OAM-based modes and the optimal SVD approach. They use a continuous propagation model for near-field communication and formulate the channel capacity using a standard formula, assuming equal power allocation (EPA) for simplicity in comparing precoding schemes.

Datasets

Not discussed in the paper. The evaluation is based on simulated channel models and configurations.

Key Insights and Ideas

  • Hardware Simplification is Key: The core insight is that by moving from continuous phase shifts (OAM) to binary phase shifts (polar Walsh functions), significant hardware complexity reduction is achievable for the precoder and combiner. This is crucial for the practical deployment of dense holographic MIMO systems.
  • Polar Walsh Functions as a Viable Alternative: The proposed polar Walsh functions offer a compelling trade-off. While they don't perfectly match the performance of OAM or SVD, their performance degradation is marginal, making them a strong candidate for scenarios where hardware cost and complexity are primary concerns.
  • Near-Field vs. Far-Field Performance: The paper highlights that the performance of different precoding schemes can vary significantly between near-field and far-field communication regimes, with continuous aperture systems showing a distinct bell-shaped capacity trend in the near-field.
  • Binary vs. Complex Precoding: Interestingly, the paper finds that switching from complex-valued OAM modes to binary-valued Walsh modes has a minimal impact on communication performance, further emphasizing the hardware benefits of the latter.

Key Results

The paper's numerical results, presented in Figure 3, compare the channel capacity of various precoding schemes under equal power allocation (EPA).

  • SVD with water-filling provides the optimal performance, as expected, achieving the highest capacity in the near-field.
  • SVD with EPA still outperforms other schemes, demonstrating the advantage of perfect CSI.
  • OAM-based modes (focused and unfocused) achieve capacities in the range of 56-74 bit/s/Hz. Focused OAM shows a slight improvement over unfocused OAM, especially in the near-field.
  • Polar Walsh functions (including radial and angular subsets) achieve capacities comparable to OAM modes, with only a marginal degradation in the near-field regime. For instance, polar Walsh functions (μ=2, ν=2) with EPA show a capacity around 60 bit/s/Hz in the near-field.
  • Radial Walsh functions exhibit substantially lower performance compared to other schemes.

Limitations and Future Work

  • Performance Gap: The authors acknowledge that the proposed polar Walsh function scheme exhibits a slight performance degradation compared to OAM-based approaches.
  • Power Allocation: The current analysis uses Equal Power Allocation (EPA) for simplicity. The authors suggest that investigating optimal power allocation strategies for the polar Walsh functions could further improve their capacity and potentially close the gap with OAM.
  • Hardware Complexity Analysis: While the paper highlights the potential for hardware simplification due to binary phases, a rigorous quantitative analysis of the actual hardware benefits and implementation complexity is suggested for future work.
  • Inter-mode Interference: The paper focuses on capacity and assumes ideal conditions. Future work could delve deeper into the impact of inter-mode interference when transmission modes are not eigenfunctions of the channel, especially in non-line-of-sight scenarios.

Summary updated

Abstract

Holographic multiple-input multiple-output (MIMO) is deemed as a promising technique beyond massive MIMO, unleashing near-field communications, localization, and sensing in the next-generation wireless networks. Semi-continuous surface with densely packed elements brings new opportunities for increased spatial degrees of freedom (DoFs) and spectrum efficiency (SE) even in the line-of-sight (LoS) condition. In this paper, we analyze holographic MIMO performance with disk-shaped large intelligent surfaces (LISs) according to different precoding designs. Beyond the well-known technique of orbital angular momentum (OAM) of radio waves, we propose a new design based on polar Walsh functions. Furthermore, we characterize the performance gap between the proposed scheme and the optimal case with singular value decomposition (SVD) alongside perfect channel state information (CSI) as well as other benchmark schemes in terms of channel capacity. It is verified that the proposed scheme marginally underperforms the OAM-based approach, while offering potential perspectives for reducing implementation complexity and expenditure.
BibTeX
@misc{vanwynsberghe2023walsh,
  author = {Charles Vanwynsberghe and Jiguang He and Chongwen Huang and Mérouane Debbah},
  title = {{Walsh Meets OAM in Holographic MIMO}},
  year = {2023},
  eprint = {2305.13868},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2305.13868},
}
Figure