![[2511.08852] DRL-Based Beam Positioning for LEO Satellite Constellations with Weighted Least Squares](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2511.08852] DRL-Based Beam Positioning for LEO Satellite Constellations with Weighted Least Squares
Summary
This paper presents a novel reinforcement learning framework for beam positioning in low Earth orbit (LEO) satellite constellations, achieving significant accuracy improvements over traditional methods.
Why It Matters
As satellite technology evolves, precise positioning is crucial for effective communication and navigation. This research introduces a method that enhances localization accuracy while reducing complexity, which is vital for the deployment of next-generation satellite networks.
Key Takeaways
- Introduces a reinforcement learning-based framework for beam positioning.
- Achieves a 99.3% reduction in mean positioning error compared to geometry-based methods.
- Utilizes uplink pilot responses for robust localization without explicit channel state information (CSI) estimation.
- Implements an augmented weighted least squares estimator for improved numerical stability.
- Demonstrates near real-time inference capabilities with a root mean square error of 0.395 m.
Electrical Engineering and Systems Science > Signal Processing arXiv:2511.08852 (eess) [Submitted on 12 Nov 2025 (v1), last revised 13 Feb 2026 (this version, v2)] Title:DRL-Based Beam Positioning for LEO Satellite Constellations with Weighted Least Squares Authors:Po-Heng Chou, Chiapin Wang, Kuan-Hao Chen, Wei-Chen Hsiao View a PDF of the paper titled DRL-Based Beam Positioning for LEO Satellite Constellations with Weighted Least Squares, by Po-Heng Chou and 3 other authors View PDF HTML (experimental) Abstract:In this paper, we propose a reinforcement learning based beam weighting framework that couples a policy network with an augmented weighted least squares (WLS) estimator for accurate and low-complexity positioning in multi-beam LEO constellations. Unlike conventional geometry or CSI-dependent approaches, the policy learns directly from uplink pilot responses and geometry features, enabling robust localization without explicit CSI estimation. An augmented WLS jointly estimates position and receiver clock bias, improving numerical stability under dynamic beam geometry. Across representative scenarios, the proposed method reduces the mean positioning error by 99.3% compared with the geometry-based baseline, achieving 0.395 m RMSE with near real-time inference. Comments: Subjects: Signal Processing (eess.SP); Machine Learning (cs.LG); Networking and Internet Architecture (cs.NI) Cite as: arXiv:2511.08852 [eess.SP] (or arXiv:2511.08852v2 [eess.SP] for this version) h...
