Source-linked AI summary
Finite-Aperture Planar Fluid Antenna Array
Zhentian Zhang, Jingyuan Xu, Kai-Kit Wong, Hao Jiang, Zaichen Zhang, Hyundong Shin
TL;DR
Planar fluid-array geometry diversity remains insufficiently understood under finite-aperture constraints, where limited physical extent shapes port-placement trade-offs. This paper develops analytical characterizations and a regularized selection method, showing predictable spacing, CRB behavior, and a precision–ambiguity trade-off.
Problem
Geometry diversity in planar fluid arrays remains insufficiently investigated, while finite apertures constrain geometric spread and the interplay among port placement, estimation accuracy, and spectral ambiguity.
Method
The paper develops a systematic analytical framework for planar finite-aperture arrays and proposes a regularized greedy selection algorithm balancing geometric precision and port-placement diversity.
Results
The minimum inter-port distance follows a Rayleigh law with O(M^-1) scaling; CRBs are governed by a 2×2 geometric inertia matrix, and determinant-optimal placement creates a precision–ambiguity trade-off.
Takeaways & Limitations
Moderate diversity regularization can balance high-precision boundary placement against lower-ambiguity interior dispersion, while the proposed greedy method outperforms uniform-grid and random baselines in geometric determinant.
Takeaways & Limitations
Pure determinant-optimal selection can concentrate ports on the aperture boundary, leaving the interior void and producing large spatial gaps that increase sidelobe levels and spectral ambiguity.
Abstract
from arXiv · showhide
Fluid antenna systems (FASs) are emerging as a reconfigurable-aperture technology that expands physical-layer design beyond fixed, rigid antenna geometries. While the \emph{fading diversity} of FASs -- which exploits spatial channel fluctuations for signal enhancement and interference avoidance -- has been widely studied, the \emph{geometry diversity} created by reconfigurable port placement remains far less understood, particularly for planar architectures under finite-aperture constraints. This paper develops a systematic analytical framework for finite-aperture planar fluid antenna arrays (FAAs). First, we derive a closed-form characterization of the minimum inter-port distance under uniform random placement over a rectangular aperture and show that it follows a Rayleigh law. Its mean scales as $\mathcal{O}(M^{-1})$, in sharp contrast to the $\mathcal{O}(M^{-2})$ behavior in the linear case in which $M$ represents the number of candidate ports, revealing a fundamentally more favorable packing geometry in two dimensions. Secondly, we establish a universal Cramér-Rao bound (CRB) for joint elevation-azimuth estimation, governed by a $2\times 2$ \emph{geometric inertia matrix} whose determinant and eigenstructure fully capture the role of port placement in estimation precision. We further prove that both the trace and determinant of this matrix are invariant to the azimuth look direction. Third, we uncover an intrinsic \emph{precision--ambiguity trade-off}: maximizing the geometric determinant to minimize the CRB drives ports toward the aperture boundary, but simultaneously increases sidelobe-induced spatial ambiguity.
I. INTRODUCTION
The paper studies geometry diversity in finite-aperture planar fluid antenna arrays, addressing planar-specific spacing, estimation, and placement challenges. It develops analytical tools centered on minimum-distance statistics and a geometric-inertia representation of joint angle estimation.
- Planar geometry diversity remains less understood than fading diversity, particularly for finite-aperture fluid arrays.
- The framework addresses three challenges: planar minimum-distance scaling, matrix-governed joint-angle CRBs, and placement-induced precision–ambiguity trade-offs.
- The minimum inter-port distance under uniform random placement follows a Rayleigh law with mean scaling O(M^-1), unlike the linear-array O(M^-2) law.
- Joint elevation–azimuth estimation is characterized by a universal closed-form CRB based on a 2×2 geometric inertia matrix.
- The planar FAA model uses M active ports within a rectangular aperture, with four corner ports fixed and remaining ports freely placed subject to a minimum distance.
III. MINIMUM RADIUS DISTRIBUTION ANALYSIS
This section derives the minimum-distance distribution for uniformly random planar port placement using collision geometry and a Poisson pair-count approximation. The resulting closed-form characterization is Rayleigh and provides scaling and calibration results for finite-aperture design.
- Uniformly random placement of M ports over a rectangular aperture is analyzed through pairwise Euclidean distances and Poisson pair counting.
- For small r relative to aperture dimensions, collision probability is governed by the fractional area of a radius-r disk within the aperture.
- The Chen–Stein approximation becomes asymptotically exact in the concentration regime r = O(M^-1√A), yielding the minimum-distance distribution.
- Proposition 1 gives closed-form CCDF, PDF, mean, and variance for R_min, with a Rayleigh PDF and scale parameter.
- The mean minimum distance scales as O(M^-1), and the result supports selecting d_min through a prescribed collision-probability threshold.
D. Comparison with the Exact Linear-Array Result
The paper contrasts planar minimum-distance analysis with the exact linear-array result. The difference reflects dimensionality and the distinct tractability of one-dimensional order statistics versus two-dimensional exclusion geometry.
- In the linear case, uniformly placed ports have an exact minimum-spacing result derived from order statistics and a simplex transformation.
- Two-dimensional pairwise distance constraints lack a natural ordering, making exact volume calculations intractable and motivating the Poisson approximation.
- The paper presents a side-by-side asymptotic comparison of planar and linear minimum inter-port distances in Table I.
IV. UNIVERSAL CRB FOR PLANAR FAA
The paper derives a universal closed-form CRB for joint elevation-azimuth estimation in finite-aperture planar arrays using a 2×2 geometric inertia matrix. Its trace and determinant are azimuth-invariant, while determinant, eigenstructure, and port placement govern precision and ambiguity.
- CRB formulation: A 2×2 geometric inertia matrix built from projected, centered port positions governs the closed-form CRB for joint elevation-azimuth estimation.The derivation treats port positions as deterministic and applies to arbitrary planar configurations.
- Geometry and precision: The CRB becomes singular when det(Lgeo)=0, which occurs when the ports are collinear after projection.When Lqr=0, the two angle estimates decouple; aligned rectangular grids satisfy this condition for the corresponding look direction.
- Azimuth invariance: The trace and determinant of Lgeo are invariant to azimuth look direction because rotating the projection axes preserves the scatter-matrix eigenvalues.Azimuth changes redistribute spread between Lqq and Lrr without changing these matrix invariants.
- CRB formulation: The Fisher information matrix separates angular factors from geometric inertia entries, generalizing the scalar geometry factor of linear arrays.The angular terms contain cos^2θ, sin^2θ, and cosθ sinθ, while Lqq, Lrr, and Lqr encode port distribution.
- Geometry and precision: Maximizing det(Lgeo) simultaneously lowers CRB(θ) and CRB(φ), while isotropic scatter minimizes the worst of the two CRBs over rotations.The determinant equals the product of the two eigenvalues of Lgeo.
- Precision–ambiguity trade-off: Corner-heavy placements maximize geometric determinant and reduce CRB but create sparse apertures with grating lobes, whereas uniform grids suppress false peaks at higher CRB.This establishes a precision–ambiguity trade-off for planar port placement.
V. ARRAY PLACEMENT ALGORITHM DESIGN
The proposed placement algorithm navigates the precision–ambiguity trade-off by selecting ports to improve the CRB-relevant geometric determinant while preserving spatial diversity.
- Algorithm objective: The algorithm maximizes the CRB-relevant geometric determinant while maintaining spatial diversity across the aperture interior.It is designed to computationally navigate the precision–ambiguity trade-off identified in the CRB analysis.
- Algorithm structure: Algorithm 1 is a regularized greedy port-selection procedure for CRB minimization.The listed implementation uses candidate generation, admissibility filtering, and running geometric sums.
A. Discrete Candidate Formulation
The discrete candidate formulation converts the continuous rectangular aperture into a regular grid and filters candidates using a minimum-distance constraint from the corner ports.
- Candidate generation: A grid with spacing Δ≤dmin discretizes the aperture into a finite candidate set.The candidate set is formed from the continuous region [0, Wx] × [0, Wy].
- Candidate filtering: A candidate is admissible only when its distance from every corner port is at least dmin.The filtered set C contains Nc admissible candidates.
B. Limitation of Pure Determinant-Optimal Selection
Pure determinant-optimal selection minimizes the CRB in isolation but concentrates ports at the aperture boundary, producing spatial gaps and increased ambiguity.
- Failure mode: Maximizing det(Lgeo) drives greedy selection toward boundary positions because they provide the largest squared deviations from the centroid.The resulting placement leaves the aperture interior void.
- Failure mode: Boundary-concentrated ports create a nonuniform virtual co-array with large gaps, increasing sidelobe levels and spectral ambiguity.The algorithmic artifact is the same precision–ambiguity trade-off identified in the CRB analysis.
C. Regularized Greedy Selection
The regularized greedy selector augments determinant-based precision optimization with a minimum-separation reward, balancing geometric precision against spatial diversity.
- C. Regularized Greedy Selection: The criterion combines geometric-determinant gain with a minimum-separation reward to disperse ports across the aperture interior.The feasible candidate set excludes locations closer than dmin to selected ports.
- C. Regularized Greedy Selection: β ≥ 0 controls the precision-diversity trade-off through the regularized selection criterion.β weights the diversity term relative to CRB-relevant determinant maximization.
- C. Regularized Greedy Selection: β = 0 yields pure D-optimal selection, with ports tending toward the aperture boundary.Boundary positions maximize squared deviations from the centroid and therefore favor determinant-based precision.
- C. Regularized Greedy Selection: β →∞ produces maximum-dispersion packing, whereas moderate β places some ports in the interior without excessively sacrificing geometric spread.Interior placement populates the co-array more uniformly than pure D-optimal selection.
D. Efficient O(1) Evaluation
The geometric determinant can be updated from running coordinate sums, while candidate-distance evaluation keeps the total greedy procedure computationally modest.
- D. Efficient O(1) Evaluation: The 2×2 geometric determinant is evaluated from running sums of coordinate squares, cross-products, and coordinate totals.These sums represent Sxx, Syy, Sxy, Sx, and Sy for the currently selected ports.
- D. Efficient O(1) Evaluation: Adding a candidate updates each running sum with a single addition.The current port count is Mk = |S|.
- D. Efficient O(1) Evaluation: The minimum-distance term costs O(Mk) per candidate, giving total complexity O((M −4)NcM).The stated total complexity remains negligible.
E. Selection of β
The dimensionless parameter β0 calibrates the relative emphasis on estimation precision and spatial diversity across aperture sizes and port counts.
- E. Selection of β: β0 is normalized using the initial four-corner set so precision and diversity terms have comparable scale.The normalization accounts for the determinant term and aperture-area-scale squared distances.
- E. Selection of β: β0 = 1 balances precision and diversity, while β0 > 1 favors spatial diversity and β0 < 1 favors estimation precision.The normalized parameter is intended to remain comparable across aperture sizes and port counts.
- E. Selection of β: The β0 trade-off is evaluated using CRB, det(Lgeo), and the number of interior ports.The figure tracks these quantities across β0 for the default configuration.
F. Computational Complexity
The proposed placement algorithm has a two-phase complexity determined by candidate-grid size and the number of ports. Its evaluation compares geometric precision, ambiguity, and spatial-diversity metrics under fixed experimental settings.
- Algorithmic complexity: Phase I runs in O(Nc), while Phase II costs O((M −4) · Nc · M) across candidate scans.The second phase iterates over the M −4 non-corner ports and evaluates up to Nc candidates with O(M) work per candidate.
- Candidate construction: The candidate grid uses spacing 0.1λ, retains all four corners, and selects the remaining M −4 ports algorithmically.Grid points violating dmin relative to corner ports are excluded before selection.
- Evaluation metrics: The study reports det(Lgeo), elevation CRB, PSL, and the number of interior ports to assess precision, ambiguity, and spatial diversity.The default evaluation uses a 2λ × 2λ square aperture with M = 25 ports and dmin = 0.2λ.
- Comparative results: The proposed greedy algorithm consistently outperforms uniform-grid and random placements in det(Lgeo), with larger gaps for larger (W, M) configurations.Increasing aperture size while proportionally scaling port count also reduces CRB across SNR levels.
- Precision–ambiguity trade-off: Increasing β0 trades lower det(Lgeo) and higher CRB for improved sidelobe suppression and more interior ports.The placement transition is sharp: free ports initially remain on the boundary, then migrate inward beyond a critical β0.