Greedy Optimization for Enhancing Quantum Key Distribution Performance in Satellite Networks

Quantum Key Distribution (QKD) is a fundamental technology to ensure secure communication between distant points on the Earth, leveraging the principles of quantum mechanics (QM) to provide an unconditionally secure method for distributing cryptographic keys. In this paper, we investigate the optimization of QKD systems in low Earth orbit (LEO) through a simulation-based approach, utilizing a Greedy Optimization Agent (GOA) that dynamically adjusts key system parameters in response to changing atmospheric and orbital conditions. The agent adjusts the beam pointing, transmission loss, and coherence factors to maximize the quantum key rate while accounting for signal degradation due to atmospheric turbulence and misalignment. The experimental results demonstrate that the optimization of the QKD system significantly outperforms the baseline system. The optimization system outperformed the baseline by 13.82 %, with an average reward of $95.80 \pm 72.63$ compared to $84.07 \pm 70.31$. In the same way, there was a 13.22 % rise in the average key rate, with the optimized system reaching $98.37 \pm 71.80$, rather than the baseline's $86.95 \pm 69.52$. In addition, the modified system significantly improved alignment by lowering the beam pointing error from 0.0456 in the baseline to 0.0170. These findings confirm that real-time dynamic adjustment is an effective strategy for improving the performance of satellitebased quantum communication systems.