Lunar Orbital VLBI Experiment (LOVEX)

• LOVEX is a space–ground VLBI system that combines a 4.2 m X-band antenna on QueQiao-2 with Earth-based telescopes to form baselines up to 380,000 km.
• It enables ultra-long-baseline astrophysics and near-field spacecraft tracking, improving brightness temperature reach and astrometric precision.
• The experiment functions both as an operational interferometer and a pathfinder for future lunar-based VLBI infrastructures and high-frequency astronomy.

Lunar Orbital VLBI Experiment (LOVEX) is a space–ground very long baseline interferometry system implemented as a scientific element of the Chinese Lunar Exploration Project within the Chang’E-7 framework. Its spaceborne component is hosted on the QueQiao-2 relay satellite, launched on 2024 March 20 and later placed into an elliptical selenocentric orbit, where a 4.2 m X-band radio telescope, a hydrogen maser frequency standard, and mission-specific VLBI data acquisition electronics form a space antenna that co-observes with Earth-based radio telescopes on baselines approaching the Earth–Moon separation. In the literature, LOVEX appears both as an operational X-band Earth–Moon-class interferometer for astronomy and near-field spacecraft tracking, and as a pathfinder that motivates broader lunar-based VLBI architectures and future lunar infrastructure (Hong et al., 22 Jul 2025, Zhao et al., 6 Jan 2026).

1. Mission identity and programmatic setting

LOVEX is a scientific component of the Chinese Lunar Exploration Project and is implemented onboard the relay satellite QueQiao-2. The system uses QueQiao-2 as a spaceborne radio telescope at X band and combines it with Earth-based radio telescopes to form baselines extending up to approximately 380,000 km. The principal scientific drivers stated for the mission are ultra–long-baseline astrophysics at X band, especially studies of compact structures in active galactic nuclei and extreme brightness temperatures, and near-field VLBI for ultra-precise orbit determination and deep-space navigation on baselines longer than the Earth’s diameter (Hong et al., 22 Jul 2025).

The present implementation is specifically an orbital X-band system. This distinction matters because adjacent literatures sometimes discuss related but non-identical concepts: an autonomous lunar surface observatory working with the terrestrial IVS network, lunar lander–orbiter radioscience using VLBI and Same Beam Interferometry, or a future lunar-based element at 230 GHz for black hole shadow work (Kurdubov et al., 2019, Gromov et al., 2015, Zhao et al., 6 Jan 2026). Those studies are directly relevant to LOVEX, but they are not equivalent to the current QueQiao-2 configuration.

LOVEX has also been characterized as the fourth operational space VLBI system worldwide. Its role is therefore both operational and methodological: it is already producing space–ground interferometric detections, while also serving as a development platform for future space-VLBI astrometry, gravitational physics experiments via precision delay modeling on long baselines, and Earth–space near-field VLBI workflows (Hong et al., 22 Jul 2025).

2. Spaceborne and ground-system architecture

The QueQiao-2 implementation combines LOVEX-specific hardware with nominal spacecraft subsystems that contribute to the VLBI function. The result is a complete space radio telescope rather than a single-purpose receiver payload (Hong et al., 22 Jul 2025).

Subsystem	Specification	Function
Antenna	4.2 m fixed-mount deployable paraboloid; aperture efficiency ≈ 35%; primary-beam FWHM ≈ 0.6°	Spaceborne collecting area for X-band VLBI
Receiver	RF coverage 8.1–9.0 GHz; dual circular polarization; maximum instantaneous bandwidth 512 MHz per polarization; in-flight $T_{\rm sys} \approx 89.3 \pm 5$ K	Low-noise X-band reception
Frequency standard	Passive H-maser with 10 MHz output; Allan deviation $\sigma_y(\tau) \approx 8\times10^{-14}$ at $\tau=100$ s and $3.6\times10^{-14}$ at $\tau=500$ s	Timing and coherence
Backend	IF 0.1–1.0 GHz; digitized up to 2048 Msps; selectable bandwidths 64, 128, 256, 512 MHz; maximum raw rate 2048 Mbps	VLBI formatting and acquisition
Storage/downlink	4 Tbit solid-state recorder; Ka-band downlink at 26 GHz using 8PSK at 500 Mbps	Buffering and return of raw VLBI data

The antenna is a 4.2 m fixed-mount deployable paraboloid optimized for X band, with approximate antenna gain at 8.4 GHz of $$G \approx \eta_a (\pi D/\lambda)^2 \approx 4.8\times10^4$$ or about 46.8 dBi. The VLBI feed is offset by 1.6° from the relay communications feed, and in-flight calibration yielded offsets in the spacecraft frame of about 0.153° and 1.774° along the X and Y axes, respectively; after calibration, the pointing precision is about 0.1° (1$\sigma$) (Hong et al., 22 Jul 2025).

The cryogenic receiver covers 8.1–9.0 GHz in dual circular polarization and supports up to 512 MHz instantaneous bandwidth per polarization. A pulse-tube cryocooler cools the LNA to $-138.2^\circ$C, with receiver noise temperature about 42.3 K and measured in-flight system temperature $T_{\rm sys} \approx 89.3 \pm 5$ K. The VLBI backend converts RF to 0.1–1.0 GHz IF, digitizes up to 2048 Msps, and records at up to 2.048 Gbps. The 4 Tbit recorder implies about 32.6 minutes of storage at the maximum rate, while a full recorder dump at 500 Mbps requires about 8000 s, or roughly 2.2 h (Hong et al., 22 Jul 2025).

The ground segment initially comprised Tianma 65 m, Kunming 40 m, and Nanshan 26 m, with Changbaishan 40 m and Shigatse 40 m joining in June 2025. Representative X-band SEFDs are approximately 50 Jy for Tianma, 275 Jy for Kunming, 345 Jy for Nanshan, 86 Jy for Changbaishan, and 72 Jy for Shigatse. In combination with the space element, whose SEFD is about $5\times10^4$ Jy, the LX–ground baselines achieve 1$\sigma_y(\tau) \approx 8\times10^{-14}$0 sensitivities of order 5–13 mJy at 512 MHz over 300 s, depending on the ground station (Hong et al., 22 Jul 2025).

Correlation is performed by a dedicated CPU+GPU correlator that supports up to 10 stations, dual polarization, bandwidth up to 512 MHz per polarization, and a $\sigma_y(\tau) \approx 8\times10^{-14}$1 ms delay search window. It generates delay models for both Earth stations and the selenocentric space antenna, performs wideband fringe searches for both far-field and near-field signals, and outputs residual delay and delay rate products for orbit-determination workflows (Hong et al., 22 Jul 2025).

3. Delay observables, baseline geometry, and calibration logic

LOVEX inherits the standard VLBI geometric scaling relations. Its angular resolution is governed by

$\sigma_y(\tau) \approx 8\times10^{-14}$2

with wavelength $\sigma_y(\tau) \approx 8\times10^{-14}$3 m at 8.4 GHz. For a maximum baseline near $\sigma_y(\tau) \approx 8\times10^{-14}$4 m, the maximum angular resolution is about $\sigma_y(\tau) \approx 8\times10^{-14}$5 rad, or approximately 19 $\sigma_y(\tau) \approx 8\times10^{-14}$6as. The core delay and fringe observables are the geometric delay

$\sigma_y(\tau) \approx 8\times10^{-14}$7

the fringe phase

$\sigma_y(\tau) \approx 8\times10^{-14}$8

and the fringe rate

$\sigma_y(\tau) \approx 8\times10^{-14}$9

For baseline sensitivity, the mission uses the usual fringe-rms relation

$\tau=100$0

with $\tau=100$1 for 2-bit quantization; for the LX–TM baseline at 512 MHz and 300 s, the quoted 1$\tau=100$2 value is about 5 mJy (Hong et al., 22 Jul 2025).

The orbit of QueQiao-2 is a selenocentric “frozen” retrograde orbit with periselene about 300 km, aposelene about 160,000 km, inclination about 118.2°, and period about 24 h. Because the baseline evolves with both lunar and terrestrial motion, projected baselines can be either very long or, for sources near the lunar orbital plane, substantially shorter by projection. The mission therefore distinguishes between geometries optimized for non-imaging detections on the longest baselines and geometries favorable for hybrid Earth–space imaging (Hong et al., 22 Jul 2025).

For near-field targets, the relevant formulation departs from the far-field plane-wave approximation. The COMPASS study, which explicitly frames its beacon methodology as support for a Lunar Orbital VLBI Experiment, writes the near-field geometric delay for lunar targets as

$\tau=100$3

where $\tau=100$4 and $\tau=100$5 are Earth-station positions and $\tau=100$6 is the spacecraft beacon position, with light-time consistency enforced. In that framework, differential or phase-referenced VLBI with simultaneous quasar calibrators cancels first-order tropospheric, ionospheric, clock, and instrumental terms, while preserving the near-field signature needed for positioning (Eubanks, 2020).

The same COMPASS framework states that coherent multi-frequency ultra-wideband beacons, interoperable with VLBA, IVS/VGOS, and future ngVLA-style networks, could provide picosecond-level interferometric phase delays during routine sessions. Its linearized position update is written as

$\tau=100$7

with $\tau=100$8, and the paper states that multi-baseline phase-referenced observations with simultaneous calibrators should enable sub-meter transverse positioning of lunar orbiters and meter-level lunar orbit determination in a few seconds per position determination (Eubanks, 2020). This suggests a direct navigation-and-timing extension to the current LOVEX architecture, although the current QueQiao-2 implementation is centered on X-band space–ground VLBI rather than on the COMPASS beacon concept.

4. Scientific objectives and demonstrated performance

LOVEX has two verified operational modes: astronomical far-field VLBI and near-field spacecraft tracking. For astrophysics, the stated objectives include studies of ultra-compact AGN structure, extreme brightness temperatures, and selected imaging experiments. Because the brightness-temperature lever arm scales as $\tau=100$9, the paper states that LOVEX improves brightness-temperature reach over Earth-only, VSOP, and RadioAstron by factors of order about 1600, about 200, and about 2.5, respectively, for fixed correlated flux. Prime targets listed include AO 0235+164, OJ 287, 3C 273, and IDV-prone blazars with compact cores (Hong et al., 22 Jul 2025).

The mission’s in-flight validation sequence established the hardware chain before science operation. Sequential activation on 28 June 2024 confirmed H-maser lock at 10 MHz, stable cryogenic operation, correct gain and current behavior, and standard VLBI data generation in all bandwidth modes. System temperature was calibrated between July and September 2024 using both the internal noise source and external cold-sky/Moon switching; the cold-sky versus Moon method, using lunar brightness temperature about 235 K, gave $3.6\times10^{-14}$0 K. Raster scans of Taurus A on 10 September and 31 October 2024 refined feed-offset calibration and yielded pointing accuracy better than 0.1°, with measured beam profiles matching pre-flight laboratory patterns (Hong et al., 22 Jul 2025).

The first in-flight astronomical detections were obtained on the blazar AO 0235+164. On 18 October 2024, using 64 MHz bandwidth from 8428 to 8492 MHz and 300 s integration, fringes were detected on all LX–ground baselines with projected baseline lengths of about 0.37 Earth diameters. On 23 January 2025, using 512 MHz bandwidth from 8108 to 8620 MHz and projected baselines of about 5.5 Earth diameters, fringes were again detected on all LX–ground baselines. These observations demonstrated robust wideband correlation on long space baselines (Hong et al., 22 Jul 2025).

The first near-field result was obtained on the Chang’E-6 orbital module at the Sun–Earth L2 point. On 18 October 2024, fringes were obtained on all baselines to the space element, and the LX–TM projected baseline was about 12 Earth diameters. The cross-spectrum showed the main carrier and two subcarriers with coherent phases. The mission paper identifies this as the first near-field VLBI detection on an Earth–space baseline and as a demonstration of the navigation and orbit-determination mode (Hong et al., 22 Jul 2025).

In later high-frequency concept work, LOVEX is treated explicitly as the first successful Moon–Earth VLBI demonstration within China’s Chang’E-7 program. That work uses LOVEX as evidence that precise lunar station ephemerides, transformation to the ICRF for correlation, long-duration Moon–Earth VLBI observations, and operational constraints such as mutual visibility and solar/lunar limb avoidance can be handled in practice (Zhao et al., 6 Jan 2026).

5. Antecedents, adjacent concepts, and modeling frameworks

Several earlier and parallel research lines define the broader technical meaning of LOVEX. One is the Russian Luna-Glob/Luna-Resource radioscience program, which described a lander radio beacon and an orbiter receiver supporting orbital Doppler measurements, VLBI interferometry, and Same Beam Interferometry. In that architecture, an accuracy of acceleration measurements in the Lander–Orbiter experiment could be about 3–10 mGal, while VLBI and SBI measurements of relative lander distances with accuracy better than millimeters were intended to support precise determination of the orbital and rotational movement of the Earth and the Moon, mass distribution and internal movements in the Moon’s interior, and checks of general relativity effects (Gromov et al., 2015). Although that paper does not use the LOVEX name explicitly, it matches core LOVEX themes: lunar geodesy, differential interferometry, and orbit-dynamics recovery.

A second antecedent is Earth–Moon VLBI modeling for a lunar observatory working with the terrestrial IVS network. That study simulated an Earth–Moon baseline up to 410,000 km and reported more than 10× improvement in source position formal errors for about half the sources away from the ecliptic, maximum 5-year lunar orbital position error reduced from 4.6 m to 0.6 m when VLBI was added to LLR, maximum physical libration error reduced from 1.3 m to 0.15 m, and PPN $3.6\times10^{-14}$1 uncertainty improved from $3.6\times10^{-14}$2 in the Earth-only “Legacy-1+2” case to $3.6\times10^{-14}$3 in the Earth–Moon case (Kurdubov et al., 2019). Those results were derived for a lunar surface element rather than an orbiter, but they define the scientific scale of what Earth–Moon-class baselines can contribute to astrometry, ephemerides, libration recovery, and relativistic delay tests.

A third line is the Chinese space-VLBI roadmap. A 2019 review summarized two mission lines—mm/sub-mm arrays and low-frequency large-aperture arrays—and described ultra-long-wavelength pathfinders already flown around and on the Moon under Chang’E-4, including Longjiang-2 in lunar orbit, the Netherlands–China Low-Frequency Explorer on the Queqiao relay satellite, and a ULW instrument on the Chang’E-4 far-side lander. The same review discussed a planned relay satellite with a roughly 4 m antenna around 2023 that could act as a Moon-orbit VLBI station with the terrestrial VLBI network to test Moon-based VLBI payload design and data processing workflows (An et al., 2019). LOVEX can be read as the operational realization of this general trajectory toward lunar-orbit VLBI.

A fourth line is methodological rather than observational. The OmniUV toolkit explicitly supports Earth fixed stations, Earth orbit stations, lunar fixed stations, lunar orbit stations, and Moon–Earth L1/L2 points, all unified in a celestial reference frame with DE-421 ephemerides and availability masks. Its Moon–Earth demonstration reported a synthesized beam major axis of about 0.028 mas and minor axis of about 0.011 mas for a Moon–Earth network, compared with 0.082 mas by 0.070 mas for an Earth-only configuration, and it is positioned as a general framework for end-to-end simulation of LOVEX-type geometries (Liu et al., 2022).

6. Limitations, misconceptions, and future directions

The most immediate limitation of the current QueQiao-2 implementation is not baseline length but mixed Earth–space uv-coverage. Because the space element orbits near the lunar orbital plane, projected baselines usually remain far from the uv-origin over a synodic month, which disfavors classical imaging for many targets. The mission therefore prioritizes non-imaging brightness-temperature constraints on the longest baselines and targeted imaging of sources near the lunar plane at favorable orbital phases (Hong et al., 22 Jul 2025). A common misconception is therefore to equate Earth–Moon baseline length with routine high-fidelity imaging; the mission papers present imaging as selective and geometry-dependent rather than universal.

Operational constraints also remain substantial. The present system has finite onboard recording and downlink capacity, Earth-side propagation errors remain on all ground paths, and precise orbit and clock modeling are essential for residual delay and delay-rate recovery. The passive H-maser supports multi-hundred-second coherent integrations in the clock-noise limit, and the correlator’s $3.6\times10^{-14}$4 ms delay window mitigates initial orbit and clock uncertainties, but continued refinement of the space baseline state vector is explicitly identified as part of the roadmap (Hong et al., 22 Jul 2025).

For navigation-oriented LOVEX extensions, the COMPASS paper identifies additional maturation steps: hardware qualification for coherent UWB beacons in space, on-orbit stability characterization, procedures for simultaneous beacon-plus-quasar observations with co-located antennas at multiple sites, and end-to-end processing pipelines tailored to near-field differential VLBI. The same paper also emphasizes practical constraints such as network availability, calibrator visibility, correlator and scheduling load, near-field modeling fidelity, and the need for power and thermal solutions compatible with $3.6\times10^{-14}$5 mW beacon operation through the lunar night (Eubanks, 2020).

The longer-term outlook connects LOVEX to lunar-based high-frequency astronomy. A 2026 study on black hole shadow detection at 230 GHz takes a LOVEX-like Moon–Earth element as a pathfinder and shows that a Moon–Earth baseline of 384,400 km yields $3.6\times10^{-14}$6as at 230 GHz. In that framework, six high-priority SMBH targets have first visibility nulls that are sampleable with lunar antennas from 5 m to 100 m depending on source flux and shadow size, but the same study emphasizes a clear coverage gap between Earth-only and Moon–Earth baselines and concludes that photon-ring science generally requires additional space telescopes to fill that gap, larger lunar antennas, wider bandwidth, and longer coherence times via frequency phase transfer (Zhao et al., 6 Jan 2026). A plausible implication is that LOVEX’s current significance is twofold: it is already a functioning Earth–space interferometer, and it is a systems-level precursor for denser lunar and cislunar VLBI arrays that would extend beyond the present X-band, single-space-element regime.