Complete Sampling of the u v uv Plane with Realistic Radio Arrays: Introducing the RULES Algorithm, with Application to 21 cm Foreground Wedge Removal
Vincent MacKay, Zhilei Xu, Ruby Byrne, Jacqueline N. Hewitt · 2025
Read Complete Sampling of the u v uv Plane with Realistic Radio Arrays: Introducing the RULES Algorithm, with Application to 21 cm Foreground Wedge Removal by Vincent MacKay, Zhilei Xu, Ruby Byrne, Jacqueline N. Hewitt (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-15296.txt.
Introduction
Vincent MacKay, Zhilei Xu, Ruby Byrne, and Jacqueline Hewitt design radio arrays that sample the measurement plane completely in a 2025 radio astronomy paper. Interferometers measure Fourier components of the sky on the so-called uv plane, and gaps in coverage distort images. Their RULES algorithm engineers layouts that leave no gaps over a chosen baseline range, with fewer antennas than current arrays.
Complete layouts matter for cosmology: bright foregrounds contaminate the faint twenty-one centimeter signal from the early Universe inside a wedge-shaped region of Fourier space. Simulated observations over an FM-adjacent frequency band, processed through an image-based power spectrum estimator, test whether complete sampling shrinks that wedge's damage. The results guide how next arrays might buy cleaner early-Universe data through geometry alone.
Catalog record
| CREATOR | Vincent MacKay, Zhilei Xu, Ruby Byrne, Jacqueline N. Hewitt |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-15296
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Complete Sampling of the u v uv Plane with Realistic Radio Arrays: Introducing the RULES Algorithm, with Application to 21 cm Foreground Wedge Removal Vincent MacKay, Zhilei Xu, Ruby Byrne, Jacqueline N. Hewitt Abstract We introduce the Radio-array u v uv Layout Engineering Strategy (RULES), an algorithm for designing radio arrays that achieve complete coverage of the u v uv plane, defined as, at minimum, regular sampling at half the observing wavelength ( λ \lambda ) along the u u and v v axes within a specified range of baseline lengths. Using RULES, we generate u v uv -complete layouts that cover the range 10 λ ≤ ‖ 𝐮 ‖ ≤ 100 λ 10\lambda\leq\norm{\mathbf{u}}\leq 100\lambda with fewer than 1000 antennas of diameter 5 λ 5\lambda , comparable to current and planned arrays. We demonstrate the effectiveness of such arrays for mitigating contamination from bright astrophysical foregrounds in 21 cm Epoch of Reionization observations—particularly in the region of Fourier space known as the foreground wedge —by simulating visibilities of foreground-like sky models over the 130–150 MHz band and processing them through an image-based power spectrum estimator. We find that with complete u v uv coverage, the wedge power is suppressed by sixteen orders of magnitude compared to an array with a compact hexagonal layout (used as a reference for a sparse u v uv coverage). In contrast, we show that an array with the same number of antennas but in a random configuration only suppresses the wedge by three orders of magnitude, despite sampling more distinct u v uv points over the same range. We address real-world challenges and find that our results are sensitive to small antenna position errors and missing baselines, while still performing equally or significantly better than random arrays in any case. We propose ways to mitigate those challenges such as a minimum redundancy requirement or tighter u v uv packing density. I Introduction In a radio interferometer, each antenna pair defines a baseline whose projected separation vector, in units of wavelength, corresponds to a sample in the so-called uv plane. If the array is planar, the visibility function—sampled at each uv coordinate—is Fourier conjugate to the sky intensity expressed in direction cosines, which reduce to angular coordinates in the flat-sky (small field) approximation ( Thompson et al. 2017 , pp. 767–781) . Because the number of baselines is finite, the sampling function is necessarily incomplete: bounded, often sparse, and potentially non-uniform, leading to artifacts in the reconstructed sky image, including point spread function (PSF) sidelobes, aliasing, and edge effects. The resulting map is called a dirty image , and the associated PSF, the dirty beam . Several strategies have been developed to mitigate the effects of incomplete uv sampling. Some techniques increase coverage by leveraging the temporal and spectral axes: Earth rotation synthesis takes advantage of the changing projection of baselines as the Earth rotates over time, while multifrequency synthesis uses the frequency dependence of baseline lengths—combined with the assumption that sources have a smooth spectral structure ( Thompson et al. 2017 , pp. 31–34, 578–579) . Deconvolution approaches seek to correct for incomplete sampling by iteratively modeling and subtracting sources to suppress PSF-induced distortions in the image. These include CLEAN ( Högbom 1974 ; Schwab 1984 ; Cotton et al. 2004 ; Cotton 2005 ) , A-projection ( Bhatnagar et al. 2008 ; Carozzi & Woan 2009 ) , forward modeling ( Bernardi et al. 2011 ) , and fast holographic deconvolution (FHD, Sullivan et al. 2012 ); for a comprehensive overview, we refer the reader to Sullivan et al. 2012 . Some arrays address the problem structurally through reconfigurable layouts: antennas can be moved between fixed “pads” to realize distinct and complementary configurations. These include linear rails (e.g., the Synthesis Telescope, Landecker et al. 2000 ), T- or Y-shaped tracks (e.g., the VLA, Thompson et al. 1980 ), or custom transporters that enable arbitrary moves (e.g., ALMA, Brown et al. 2004 ). Alternative array designs aim to maximize distinct uv samples from the outset: configurations with offset sub-arrays have been proposed for this purpose ( Dillon & Parsons 2016 ) , as have layouts inspired by Golomb rulers—mathematical constructs in which all pairwise differences between elements are distinct ( Biraud et al. 1974 ; Thompson et al. 2017 ; Parsons et al. 2012a ; Ebrahimi & Gazor 2023 ; Lazko & Lazko 2023 )—sometimes obtained via algorithms for optimal antenna placement ( Keto 1997 ; Boone 2001 ; Boone 2002 ; Cohanim et al. 2004 ; Murray & Trott 2018 ) . Meanwhile, large arrays increasingly adopt random or pseudo-random layouts to achieve relatively uniform uv coverage (e.g., MeerKAT, Booth et al. 2009 ; MWA, Lonsdale et al. 2009 ; DSA-2000, Hallinan et al. 2019 ; SKA, Weltman et al. 2020 ). One specific challenge that follows from incomplete uv sampling, and that is particularly severe in 21 cm cosmology, is spectral leakage of bright foregrounds in power spectrum estimates. This contamination is most important in the so-called foreground wedge , a region of the two-dimensional power spectrum at low k_{\parallel} , where the instrument’s intrinsic chromaticity and imperfect calibration causes foregrounds to spill into the cosmological signal window (see section IV , and Bowman et al. 2009 ; Datta et al. 2010 ; Morales et al. 2012 ; Parsons et al. 2012b ). The wedge not only reduces sensitivity to the 21 cm signal, but hinders cross-correlation with other surveys, prevents most imaging-based analyses ( Pober et al. 2014 ; Beardsley et al. 2015 ; Seo & Hirata 2016 ; Cohn et al. 2016 ; Cox et al. 2022 ; Gagnon-Hartman et al. 2024 ) , notably one-point statistics ( Kittiwisit et al. 2018 ; Kim et al. 2025 ) , and affects calibration negatively ( Barry et al. 2016 ; Ewall-Wice et al. 2017 ; Byrne et al. 2019 ) . While many existing methods to tackle imperfect uv sampling, such as those enumerated above, are effective at improving source localization or imaging fidelity, only some are equipped to address the foreground wedge. Pseudo-random layouts, for instance, are sometimes adopted to reduce leakage by lowering redundancy and spreading out uv coverage, but their performance is not guaranteed; they may still leave gaps or other artifacts caused by uneven sampling, and their effectiveness is difficult to predict and control. Techniques based on wedge subtraction also exist ( Liu & Tegmark 2012 ; Paciga et al. 2013 ; Liu et al. 2014 ; Mertens et al. 2018 ; Cox et al. 2024 ) , but have yet reached levels required by 21 cm science, and are often constrained by the limitations of the underlying uv sampling (see Liu & Shaw 2020 for a review). Given the fact that foregrounds and calibration systematics dominate the error budget on most baselines in current and planned 21 cm arrays—they are not yet noise limited—and that nearly all of the strategies for improved uv -sampling listed previously were developed for arrays with \lesssim 100 antennas, it is relevant to explore new design-based approaches suited to the emerging \mathcal{O}(10^{3}) -antenna era ( Vanderlinde et al. 2019 ; Hallinan