Single-Channel Lunar Penetrating Radar for Rover Localization via Revisit-Aware Amplitude-Phase Learning

GNSS-denied lunar rover localization relies heavily on surface vision, which degrades under extreme solar angles, featureless regolith, and permanent shadows. Exploiting stable, visually invariant subsurface features via single-channel Lunar Penetrating Radar (LPR) offers a promising auxiliary means, yet repurposing in-situ geophysical probing remains unexplored, because revisit benchmarks are scarce and single-channel echoes are strongly attenuated. Here we establish the first revisit-aware localization benchmark from Chang’E-4 Yutu-2 500 MHz LPR observations, keeping non-revisited training/validation traverses strictly disjoint from the physically revisited scenes used for evaluation. By disentangling single-channel sweeps into energy envelopes and continuous unit-circle phase manifolds with structure-preserving aggregation, our early-fusion representation captures high-contrast scatterers and faint stratigraphic horizons. Across lunar and terrestrial radar datasets, our approach reaches 78.05% top-1 recall on lunar revisits, outperforming raw-echo (70.73%) and the strongest learned baseline (73.17%). Existing single-channel LPR payloads can thus support localization without extra hardware in visually degraded scenarios.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-20
DOI
https://doi.org/10.21203/rs.3.rs-11642499/v1
Primary Topic
Geophysical Methods and Applications
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Single-Channel Lunar Penetrating Radar for Rover Localization via Revisit-Aware Amplitude-Phase Learning

Pengyu Zhang, Chunyu Ding, X. Huang, Jin Tian et al.
Research Square
Geophysical Methods and Applications
preprint

Single-Channel Lunar Penetrating Radar for Rover Localization via Revisit-Aware Amplitude-Phase Learning

Pengyu Zhang, Chunyu Ding, X. Huang, Jin Tian, Xiaohang Qiu, Yuwei Chen
preprint en

Abstract

GNSS-denied lunar rover localization relies heavily on surface vision, which degrades under extreme solar angles, featureless regolith, and permanent shadows. Exploiting stable, visually invariant subsurface features via single-channel Lunar Penetrating Radar (LPR) offers a promising auxiliary means, yet repurposing in-situ geophysical probing remains unexplored, because revisit benchmarks are scarce and single-channel echoes are strongly attenuated. Here we establish the first revisit-aware localization benchmark from Chang’E-4 Yutu-2 500 MHz LPR observations, keeping non-revisited training/validation traverses strictly disjoint from the physically revisited scenes used for evaluation. By disentangling single-channel sweeps into energy envelopes and continuous unit-circle phase manifolds with structure-preserving aggregation, our early-fusion representation captures high-contrast scatterers and faint stratigraphic horizons. Across lunar and terrestrial radar datasets, our approach reaches 78.05% top-1 recall on lunar revisits, outperforming raw-echo (70.73%) and the strongest learned baseline (73.17%). Existing single-channel LPR payloads can thus support localization without extra hardware in visually degraded scenarios.

Research Square
National University of Defense Technology (CN), Shenzhen University (CN)
National Natural Science Foundation of China, Natural Science Foundation of Guangdong Province, National University of Defense Technology, National Astronomical Observatories, Chinese Academy of Sciences
Geophysical Methods and Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon
Single-Channel Lunar Penetrating Radar for Rover Localization via Revisit-Aware Amplitude-Phase Learning — Pengyu Zhang, Chunyu Ding, et al. · Research Square (2026) | TGRS Research Map | TGRS