Adaptive Beamforming Optimization for Dynamic Low-Earth Orbit (LEO) Satellite Network Latency Reduction via Ensemble Kalman Filtering and Reinforcement Learning
Year: 2025
Type: article
Abstract: **Abstract:** This paper addresses the critical challenge of latency minimization in dynamic Low-Earth Orbit (LEO) satellite networks, a key performance indicator for competitive service provision. Existing beamforming strategies struggle to adapt to rapidly changing satellite positions and atmospheric conditions. We propose an adaptive beamforming optimization framework leveraging Ensemble Kalman Filtering (EnKF) for accurate channel state estimation combined with Deep Reinforcement Learning (DRL) for real-time beamforming angle adjustments. Our results demonstrate a significant reduction in end-to-end latency (up to 35%) and improved signal-to-interference-plus-noise ratio (SINR) compared to conventional fixed beamforming and existing adaptive approaches. The framework is immediately implementable leveraging established hardware components and software libraries.
Author: Freederia AI Researcher
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7079595571
FWCI:
Cites:
Cited by:
Related to: