Drone Beam Mapping of the TONE Radio Dish Array

TL;DR

This study demonstrates a drone-based calibration system to produce high-resolution 2D beam maps for 400–800 MHz observations of the TONE CHIME/FRB outrigger telescope. Using a switched broadband calibration source mounted on a DJI Matrice 600 Pro, the authors map four TONE dishes in both co- and cross-polarizations, validate results against CST simulations, and quantify polarization leakage and RFI contributions. The workflow integrates background subtraction, precise timing alignment, 2D Gaussian beam fitting, and cross-flight normalization to yield coherent coadded beam maps, revealing robust main-beam and sidelobe structures as well as polarization-systematics that affect high-frequency performance. The results underscore the potential and challenges of drone-based beam mapping for 21 cm cosmology, including the need to account for transmitter patterns, validate far-field coverage, and mitigate RFI for accurate instrumental calibration.

Abstract

Drone-based beam measurements are a promising avenue to tackle the critical challenge of calibration for 21 cm cosmology telescopes. In this paper, we introduce a new drone-based calibration system for 400-800 MHz radio observatories, describing its instrumentation and first deployment. We discuss measurements of the TONE array, a CHIME/FRB outrigger pathfinder, and present results, including full 2D high spatial resolution beam maps in both co- and cross-polarization, as well as comparisons to simulations. The polarized beam maps cover a 70 degree by 70 degree grid, capturing the first two sidelobes and measuring the TONE main beam and first sidelobe with 7-9% statistical errors. We investigate polarization angle alignment with frequency, finding significant polarization leakage in the TONE antennas at frequencies above 600 MHz, and a polarization axis rotation with frequency. We describe statistical and systematic errors, as well as measurements of radio frequency interference from the drone and equipment. Our drone system is the first to incorporate a broad-band switched calibration source in the drone payload, enabling background subtraction and direct measurements of the RFI emitted by the drone. The results presented are the first drone-based 2D measurements of cross-polar beam structure and of polarization alignment of an array. The high frequency and spatial resolution achieved with this system have revealed the rich structure of the beam of each antenna, and enabled comparisons between individual dishes and to electromagnetic simulations.

Figures (19)

• Figure 1: (Upper) A rendering of the Matrice 600 Pro, and a photograph of the payload with components indicated (described in more detail in Section \ref{['sec:drone']}). The payload consists of the calibrator source, a GPS to provide a pulse synchronized to PPS, batteries for the source and components, and the transmission antenna. (Lower) Schematic of transmission signal RF chain.
• Figure 2: TONE at Green Bank Radio Observatory. The 8 active dishes are marked with X's, color-coded by whether the dish is instrumented with a HIRAX or CHIME feed. The four dishes used for drone beam mapping are marked with a yellow circle. The drone take off location and 'Dish 0' location (which served as the coordinate system origin) are also marked.
• Figure 3: Each panel shows the beam measured from each of the flights used in this analysis for a single dish and frequency after the per-flight processing performed in Section \ref{['sec:processing']}. The cardinal direction 'North' is vertical; some flights were oriented North-South and others East-West, and color indicates beam amplitude. The polarization during the flight was either aligned North-South ('N-pol') or East-West ('E-pol'). The smaller grids target a measurement of the main beam, the larger grids target a measurement of sidelobes. Additional detail on the flights is given in Table \ref{['tab:flights']}.
• Figure 4: A single transit of the drone through near the center of the beam for Dish 0, North-polarized input from flight 625 at frequency 448 MHz. Indicated by color are points in the flight where the background fitting found the signal was on (black), off (purple), or partially on and partially off (peach). Only 'On' data is used in the following analysis after background subtraction.
• Figure 5: Beam power vs angle in a single 1D cut through the beam at the origin for frequency 448 MHz and a North-polarized input. (Left) shows a 1D cut through the North-South direction and (Right) shows a 1D cut through the East-West direction. Each flight is shown as a different color after the normalization correction procedure described in Section \ref{['sec:ampcorr']}. As described, the amplitude correction procedure matches the average signal power in the first sidelobe region. As a result, the main beam flights have good signal in the main beam and the sidelobes, while the brighter sidelobe flights will look flat within the main beam where the signal was saturated.