Abstract:The four-intensity protocol for measurement-device-independent (MDI) quantum key distribution (QKD) is renowned for its excellent performance and extensive experimental implementation. To enhance this protocol, a machine learning-driven rapid parameter optimization method is developed. This initial step involved a speed-up technique that quickly pinpoints the worst-case scenarios with minimal data points during the optimization phase. This is followed by a detailed scan in the key rate calculation phase, streamlining data collection to fit machine learning timelines effectively. Several machine learning models are assessed-Generalized Linear Models (GLM), k-Nearest Neighbors (KNN), Decision Trees (DT), Random Forests (RF), XGBoost (XGB), and Multilayer Perceptron (MLP)-with a focus on predictive accuracy, efficiency, and robustness. RF and MLP were particularly noteworthy for their superior accuracy and robustness, respectively. This optimized approach significantly speeds up computation, enabling complex calculations to be performed in microseconds on standard personal computers, while still achieving high key rates with limited data. Such advancements are crucial for deploying QKD under dynamic conditions, such as in fluctuating fiber-optic networks and satellite communications. To advance practical quantum communication, reducing parameter optimization time at end devices is essential. The four-intensity MDI-QKD is widely used for its high performance, but its computational load for parameter optimization is too burdensome for small terminals. This study introduces a machine learning approach for swiftly predicting and optimizing parameters for the four-intensity MDI-QKD, significantly boosting its practical deployment. image

Machine Learning with Neural Networks for Parameter Optimization in Twin-Field Quantum Key Distribution

Application of Machine Learning in Optimal Allocation of Quantum Communication Resources

Optimization Parameter Prediction-Based XGBoost of TF-QKD

Parameter Optimization in Decoy-State Phase-Matching Quantum Key Distribution

Parameter Optimization and Real-Time Calibration of a Measurement-Device-independent Quantum Key Distribution Network Based on a Back Propagation Artificial Neural Network

Highly Efficient Twin-Field Quantum Key Distribution with Neural Networks

Practical twin-field quantum key distribution parameter optimization based on quantum annealing algorithm

Comparative Study of Fast Optimization Method for Four-Intensity Measurement-Device-Independent Quantum Key Distribution Through Machine Learning

Improving Parameter Optimization in Decoy-State Quantum Key Distribution

Implementation of Machine Learning in Quantum Key Distributions

Parameter Optimization in Satellite-Based Measurement-Device-independent Quantum Key Distribution

Neural network-based prediction of the secret-key rate of quantum key distribution

Optimized Protocol for Twin-Field Quantum Key Distribution

Parameter Estimation of Continuous Variable Quantum Key Distribution System Via Artificial Neural Networks

Optimized multi-head self-attention and gated-dilated convolutional neural network for quantum key distribution and error rate reduction

Quantum Algorithms for Feedforward Neural Networks

Optimization of Quantum Key Distribution Parameters Based on Random Forest

Parameter optimization of SQCC-CVQKD based on genetic algorithm in the terahertz band

Simple Method for Asymmetric Twin-Field Quantum Key Distribution

Parameters Optimization of Decoy-State Phase-Matching Quantum Key Distribution Based on the Nature-inspired Algorithms

Neural Network-Based Prediction for Secret Key Rate of Underwater Continuous-Variable Quantum Key Distribution through a Seawater Channel