[Paper Review] Holographic Metasurface Antennas for Uplink Massive MIMO Systems

Insang Yoo, David R. Smith|arXiv (Cornell University)|Aug 27, 2021

Antenna Design and Analysis37 references4 citations

TL;DR

This paper proposes a low-complexity, low-cost uplink massive MIMO system using an array of rectangular waveguide-fed holographic metasurface antennas at 3.5 GHz. By leveraging subwavelength metamaterial elements to generate multiple fan beams with high azimuthal resolution, the system achieves a sum capacity of 44.75 bits/s/Hz—close to the Rayleigh channel's 46.34 bits/s/Hz—while minimizing RF chain requirements through passive beamforming.

ABSTRACT

We propose an uplink massive MIMO system using an array of holographic metasurfaces as a sector antenna. The antenna consists of a set of rectangular waveguide-fed metasurfaces combined along the elevation direction into a planar aperture, each with subwavelength-sized metamaterial elements as radiators. The metamaterial radiators are designed such that the waveguide-fed metasurface implements a holographic solution for the guided (or reference) mode, generating a fan beam towards a prescribed direction, thereby forming a multibeam antenna system. We demonstrate that a narrowband uplink massive MIMO system using the metasurfaces can achieve the sum capacity close to that offered by the Rayleigh channel at 3.5 GHz. We show that metasurfaces supporting multiple fan beams can achieve high spatial resolution in the azimuth directions in sub-6 GHz channels, and thereby form uncorrelated MIMO channels between the base station and users. Also, the proposed metasurface antenna is structurally simple, low-cost, and efficient, and thus is suitable to alleviate RF hardware issues common to massive MIMO systems equipped with a large antenna system.

Motivation & Objective

• To address the high RF hardware complexity and cost of conventional massive MIMO systems with large antenna arrays.
• To exploit the spatial properties of sub-6 GHz MIMO channels, where elevation spread is small, to simplify beamforming design.
• To develop a structurally simple, low-profile, and efficient radiative platform using waveguide-fed metasurfaces for massive MIMO.
• To demonstrate that metasurface-based beamforming can achieve sum capacity close to the theoretical Rayleigh channel limit.
• To enable passive, low-power beamforming without requiring per-element RF chains, reducing system cost and power consumption.

Proposed method

• Designing rectangular waveguide-fed metasurfaces with subwavelength metamaterial radiators to implement holographic beamforming for generating fan beams.
• Using the coupled dipole method (CDM) to model and optimize the radiation patterns of individual metasurfaces.
• Arranging multiple metasurfaces along the elevation direction to form a planar aperture with multibeam capability.
• Employing holographic principles to control phase shifts across the aperture, enabling directive beam steering without active RF components.
• Simulating the system using a statistical MIMO channel model with varying angular spreads (σᵘ, σᵇˢ) from 5° to 60°.
• Evaluating system performance via sum capacity and condition number of the MIMO channel matrix, comparing results to the Rayleigh fading channel benchmark.

Experimental results

Research questions

• RQ1 Can a holographic metasurface-based antenna array achieve sum capacity close to the Rayleigh channel in an uplink massive MIMO system at 3.5 GHz?
• RQ2 To what extent does the spatial resolution of the metasurface array reduce spatial correlation in the MIMO channel?
• RQ3 How does the aperture size (number of metamaterial elements) affect the achievable sum capacity and channel condition number?
• RQ4 How does the angular spread of user and base station clusters influence the performance of the metasurface-based MIMO system?
• RQ5 Can passive, low-power metasurfaces replace conventional RF-intensive beamforming architectures in massive MIMO systems?

Key findings

• The proposed metasurface-based massive MIMO system achieves a mean sum capacity of 44.75 bits/s/Hz, which is very close to the Rayleigh channel's 46.34 bits/s/Hz.
• The system's mean condition number reaches a minimum of 8.15 for angular spreads ≥30°, indicating low spatial correlation and favorable MIMO channel quality.
• Increasing the number of metamaterial elements (Nₘ) from 10 to 45 improves sum capacity, with saturation observed at Nₘ ≥ 30, indicating sufficient spatial resolution is achieved.
• For fixed Nₘ, increasing angular spread (σᵘ = σᵇˢ) reduces channel correlation, improving performance due to better excitation of diverse propagation paths.
• The condition number of the metasurface-based system (κₘₛ/κᵣₐyₗₑᵧ ≈ 1.8) is comparable to that of the Rayleigh channel, confirming the formation of uncorrelated subchannels.
• The system maintains high performance with a passive, low-cost, and low-profile design, eliminating the need for per-element RF chains and reducing power consumption.