Analysis of Machine Learning Techniques for Information Reconciliation in CV-QKD Systems: Challenges in Practical Scenarios

This paper reviews the challenges associated with implementing InformationReconciliation (IR) techniques in Continuous Variable Quantum Key Distribution (CVQKD) systems, focusing on integrating Machine Learning techniques and embeddedtechnologies such as Field-Programmable Gate Arrays (FPGAs). While CV-QKDsystems are compatible with existing telecommunications infrastructure and capableof high key generation rates, they face significant computational challenges in longdistance communication due to large decoding matrices. This study explores machinelearning approaches to reduce computational complexity and improve post-keyprocessing times, aiming to identify challenges and gaps that can guide future researchin enhancing the efficiency of CV-QKD systems.