Saved in:
Bibliographic Details
Main Authors: Chou, Po-Heng, Wang, Chiapin, Chen, Kuan-Hao, Hsiao, Wei-Chen
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.08852
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912969049243648
author Chou, Po-Heng
Wang, Chiapin
Chen, Kuan-Hao
Hsiao, Wei-Chen
author_facet Chou, Po-Heng
Wang, Chiapin
Chen, Kuan-Hao
Hsiao, Wei-Chen
contents This paper investigates a lightweight deep reinforcement learning (DRL)-assisted weighting framework for CSI-free multi-satellite positioning in LEO constellations, where each visible satellite provides one serving beam (one pilot response) per epoch. A discrete-action Deep Q-Network (DQN) learns satellite weights directly from received pilot measurements and geometric features, while an augmented weighted least squares (WLS) estimator provides physics-consistent localization and jointly estimates the receiver clock bias. The proposed hybrid design targets an accuracy-runtime trade-off rather than absolute supervised optimality. In a representative 2-D setting with 10 visible satellites, the proposed approach achieves sub-meter accuracy (0.395m RMSE) with low computational overhead, supporting practical deployment for resource-constrained LEO payloads.
format Preprint
id arxiv_https___arxiv_org_abs_2511_08852
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DRL-Based Beam Positioning for LEO Satellite Constellations with Weighted Least Squares
Chou, Po-Heng
Wang, Chiapin
Chen, Kuan-Hao
Hsiao, Wei-Chen
Signal Processing
Machine Learning
Networking and Internet Architecture
This paper investigates a lightweight deep reinforcement learning (DRL)-assisted weighting framework for CSI-free multi-satellite positioning in LEO constellations, where each visible satellite provides one serving beam (one pilot response) per epoch. A discrete-action Deep Q-Network (DQN) learns satellite weights directly from received pilot measurements and geometric features, while an augmented weighted least squares (WLS) estimator provides physics-consistent localization and jointly estimates the receiver clock bias. The proposed hybrid design targets an accuracy-runtime trade-off rather than absolute supervised optimality. In a representative 2-D setting with 10 visible satellites, the proposed approach achieves sub-meter accuracy (0.395m RMSE) with low computational overhead, supporting practical deployment for resource-constrained LEO payloads.
title DRL-Based Beam Positioning for LEO Satellite Constellations with Weighted Least Squares
topic Signal Processing
Machine Learning
Networking and Internet Architecture
url https://arxiv.org/abs/2511.08852