Array Geometry-Centric Axial Sidelobe Interference Analysis for Near-Field Multi-User MIMO

This paper investigates the impact of array geometry on axial sidelobe interference in near-field multi-user MIMO systems, deriving closed-form gain expressions to demonstrate that uniform planar arrays outperform other configurations by achieving the lowest sidelobe levels and highest sum rates.

The Gist

Here is an explanation of the paper using simple language and everyday analogies.

The Big Picture: From Flashlights to Spotlights

Imagine you are trying to talk to a group of friends in a large, dark park.

• The Old Way (Far-Field): In the past, wireless signals were like a flashlight. You could point the beam left, right, up, or down, but the beam stayed the same width no matter how far away your friend was. If you wanted to talk to two people standing at different distances, you couldn't do it easily because the beam would hit both of them at once, causing a mess of overlapping voices (interference).
• The New Way (Near-Field): With the new "Ultra-Massive" antennas described in this paper, the signal acts more like a high-tech spotlight. You can focus the light not just on a specific direction, but on a specific distance. You can shine a tight beam on a friend 10 meters away and a different tight beam on a friend 20 meters away, even if they are standing in the same direction. This is called Near-Field Beamfocusing.

The Problem: The "Leaky" Spotlight

The problem with these new spotlights is that they aren't perfect. Just like a real spotlight has a bright center (the main beam) but also faint, annoying rings of light around it (sidelobes), these wireless beams have "leaks."

• The Main Lobe: The bright center where your friend is standing.
• The Sidelobes: The faint rings of light leaking out to the sides.

In the old days, these leaks only happened to the left and right (lateral). But in this new "Near-Field" world, the leaks also happen forward and backward (axial).

Imagine you are trying to talk to a friend 10 meters away. The "leak" from your beam might accidentally hit a friend standing 12 meters away. This is called Axial Sidelobe Interference. It's like trying to whisper to someone while a noisy neighbor is standing just behind them, and your voice is leaking over to the neighbor.

The Experiment: Shaping the Antenna

The researchers asked a simple question: "Does the shape of the antenna array change how 'leaky' these beams are?"

They tested four different shapes, like arranging people holding flashlights in different formations:

1. The Line (ULA): Everyone stands in a single straight line.
2. The Square (USA): Everyone stands in a perfect square grid.
3. The Circle (UCA): Everyone stands in a single ring.
4. The Concentric Circles (UCCA): Everyone stands in multiple rings, like a target or a bullseye.

The Results: The Square Wins

The researchers did some heavy math and computer simulations to see which shape created the "cleanest" beam with the least amount of leakage.

• The Loser (The Circle): The single ring (UCA) was the leakiest. It had the highest "noise" around the main beam. It's like trying to focus a spotlight with a ring of people; the light scatters too much forward and backward.
• The Middle Ground (The Line & The Bullseye): The straight line and the bullseye pattern were better, but not perfect.
• The Winner (The Square): The Uniform Square Array (USA) was the clear champion.

Why the Square?
Think of the square array like a dense, solid wall of people holding flashlights. Because the square fills the space so evenly, it cancels out the "leaks" much better than the other shapes.

• The Math: The square array reduced the interference (the "noise") by a massive margin. While the circle had a "noise level" of about -8 dB, the square dropped it to -17.6 dB. In the world of signals, that's a huge difference—it means the square array is much quieter and cleaner.

The Real-World Impact: Talking to More People

Why does this matter? Because less leakage means less interference.

If you have a square antenna array:

1. You can pack more people into the same space.
2. You can talk to people at different distances without your voice leaking to the wrong person.
3. Result: The total speed of the internet (Sumrate) goes up significantly.

The paper concludes that if we want the fastest, most efficient 6G networks in the future, we should build our antennas in square grids rather than circles or lines. It's the best shape for focusing energy exactly where it's needed and keeping the "noise" out.

Summary Analogy

Imagine you are pouring water (data) into cups (users) using a hose (antenna).

• Old antennas were like a garden hose that sprayed everywhere.
• New antennas are like a precision nozzle that can hit a cup 5 feet away and another cup 10 feet away.
• The Shape Issue: If the nozzle is shaped like a ring, water splashes everywhere (high interference). If the nozzle is shaped like a solid square block, the water stream is tight, clean, and hits only the intended cups.

The Paper's Verdict: Use the Square nozzle. It's the cleanest, most efficient way to deliver data in the future.

Technical Summary

Here is a detailed technical summary of the paper "Array Geometry-Centric Axial Sidelobe Interference Analysis for Near-Field Multi-User MIMO."

1. Problem Statement

As wireless communication systems evolve toward Ultra-Massive MIMO (UM-MIMO) operating at high-frequency bands (e.g., mmWave), the radiative near-field (NF) region expands significantly. Unlike far-field (FF) systems where beams are steered only in angular directions, NF systems utilize beamfocusing, concentrating energy in both angle and range (depth).

While NF beamfocusing enables multiplexing users in both domains, it introduces a specific interference challenge: Axial Sidelobes (ASLs).

• In FF, interference primarily comes from lateral sidelobes (LSLs) in the angular domain.
• In NF, beams exhibit ASLs along the range dimension. When multiple users are separated by distances smaller than the mainlobe resolution, or even when separated by larger distances, these ASLs cause significant inter-user interference, degrading the achievable sumrate.
• The Gap: While prior research has extensively analyzed how array geometry affects beamdepth (mainlobe resolution), the behavior of ASLs across different array geometries (Linear, Rectangular, Circular, Concentric Circular) remains unexplored.

2. Methodology

The authors propose a systematic analysis to characterize ASLs and determine the optimal array geometry for minimizing them.

• System Model: A Multi-User MIMO (MU-MIMO) downlink scenario where a Base Station (BS) with $N$ isotropic antennas serves $K$ single-antenna users. The channel is modeled using a Line-of-Sight (LoS) NF array response vector based on spherical wavefronts.
• Metric Definition:
  • Axial Beam Pattern: Defined as the normalized cross-correlation between the desired user's channel and an interfering user's channel in the range domain.
  • Peak Sidelobe Level (PSLL): The ratio of the strongest sidelobe to the mainlobe, indicating worst-case interference susceptibility.
  • Integrated Sidelobe Level (ISLL): The total sidelobe power relative to the mainlobe, indicating average interference.
• Analytical Derivation:
  • The authors derived closed-form expressions for the array gain in the range domain for four specific geometries:
    1. Uniform Linear Array (ULA)
    2. Uniform Rectangular Array (URA) (specifically analyzing the Uniform Square Array, USA)
    3. Uniform Circular Array (UCA)
    4. Uniform Concentric Circular Array (UCCA)
  • Mathematical Tools:
    • For ULA and URA: Used second-order Taylor expansion of the distance term to approximate the phase, leading to Fresnel integrals ($C(\gamma)$ and $S(\gamma)$).
    • For UCA: Utilized rotational symmetry and trigonometric identities to approximate the summation as an integral, resulting in the zeroth-order Bessel function ($J_0$).
    • For UCCA: Modeled as concentric rings, approximating the summation over rings as an integral to derive a sinc-like function.
• Simulation: Numerical simulations were conducted with a fixed aperture length ($D=1.26$ m) at 15 GHz to ensure a fair comparison of geometries, varying the number of elements ($N$) to maintain the fixed aperture.

3. Key Contributions

1. First Characterization of ASLs by Geometry: The paper is the first to analytically derive and compare the axial sidelobe levels of ULA, URA, UCA, and UCCA in the near-field region.
2. Closed-Form Expressions: The authors provide exact mathematical formulations for the axial beam patterns of these geometries, linking them to Fresnel integrals and Bessel functions.
3. Identification of Optimal Geometry: The analysis reveals a counter-intuitive finding regarding the trade-off between beamdepth and sidelobe levels:
   • While elongated arrays (ULA) minimize beamdepth (narrowest mainlobe), they maximize sidelobes.
   • Uniform Square Arrays (USA) yield the lowest Axial Sidelobes, despite having a wider beamdepth compared to ULAs.
4. Quantitative Comparison: The study provides specific PSLL values for each geometry, establishing a clear hierarchy of performance.

4. Key Results

The theoretical analysis and numerical simulations yielded the following comparative results (assuming boresight focusing and fixed aperture):

• PSLL Performance (Range Domain):
  • USA (Uniform Square Array): −17.6 dB (Best performance).
  • UCCA (Uniform Concentric Circular Array): −13.4 dB.
  • ULA (Uniform Linear Array): −8.7 dB.
  • UCA (Uniform Circular Array): −7.9 dB (Worst performance).
• ISLL Performance: The USA also achieved the lowest Integrated Sidelobe Level (−12.1 dB), confirming superior average interference suppression.
• Angular vs. Range: Interestingly, while the USA is best in the range domain, the UCCA performs best in the angular domain (PSLL of −17.6 dB). However, since the paper focuses on NF range interference, the USA is the superior choice for the specific problem of ASLs.
• Sumrate Impact: Monte Carlo simulations demonstrated that the USA achieves the highest achievable sumrate. This is directly correlated to its ability to suppress ASLs, thereby minimizing inter-user interference.
• Scalability: The PSLL is shown to be invariant to the number of antenna elements ($N$); increasing $N$ does not reduce the PSLL, though it does reduce the ISLL.

5. Significance and Conclusion

This work provides critical design guidelines for future 6G and ultra-massive MIMO systems operating in the near-field.

• Design Implication: For systems prioritizing high spectral efficiency in the near-field (where users are multiplexed by range), Uniform Square Arrays (USA) are the optimal geometry choice, outperforming the traditionally popular ULA and UCA.
• Interference Management: The study highlights that suppressing axial sidelobes is distinct from suppressing lateral sidelobes. A geometry that minimizes beamwidth (ULA) may actually exacerbate range-domain interference.
• Future Directions: The authors suggest that windowing methods and flexible antenna arrays could be explored in future work to further suppress sidelobes in both axial and lateral dimensions simultaneously.

In summary, the paper establishes that array geometry is a primary lever for controlling near-field axial interference, and a square aperture configuration offers the most robust performance for multi-user near-field communications.