arXiv:1812.08388 · Journal of the Optical Society of America B
Parameter optimization and real-time calibration of a measurement-device-independent quantum key distribution network based on a back propagation artificial neural network
Selection of parameters (e.g., the probability of choosing an X-basis or Z-basis, the intensity of signal state and decoy state, etc.) and system calibrating are more challenging when the number of users of a measurement-device-independent quantum key distribution (MDI-QKD) network increases. At present, optimization algorithms are usually employed when searching for the best parameters. This method can find the optimized parameters accurately, but it may take a lot of time and hardware resources. This is a big problem in a large-scale MDI-QKD network. Here, we present, to the best of our knowledge, a new method, using a back propagation artificial neural network (BPNN) to predict, rather than search, the optimized parameters. Compared to optimization algorithms, our BPNN is faster and more lightweight, and it can save system resources. Another big problem brought by large-scale MDI-QKD networks is system recalibration. BPNN can support this work in real time, and it only needs to use some discarded data generated from the communication process, rather than adding additional devices or scanning the system.
Selection of parameters (e.g., the probability of choosing an X-basis or Z-basis, the intensity of signal state and decoy state, etc.) and system calibrating are more challenging when the number of users of a measurement-device-independent quantum key distribution (MDI-QKD) network increases. At present, optimization algorithms are usually employed when searching for the best parameters. This method can find the optimized parameters accurately, but it may take a lot of time and hardware resources. This is a big problem in a large-scale MDI-QKD network. Here, we present, to the best of our knowledge, a new method, using a back propagation artificial neural network (BPNN) to predict, rather than search, the optimized parameters. Compared to optimization algorithms, our BPNN is faster and more lightweight, and it can save system resources. Another big problem brought by large-scale MDI-QKD networks is system recalibration. BPNN can support this work in real time, and it only needs to use some discarded data generated from the communication process, rather than adding additional devices or scanning the system.
Abstract The satellite-based measurement-device-independent quantum key distribution can promote the realization of quantum communication networks. Under the condition of the limited data set, it is necessary to optimize all parameters. For communication networks, real-time prediction and optimization are also indispensable. With the development of machine learning, cross-combination with machine learning has also become the mainstream of parameter optimization in various disciplines. This paper discusses the asymmetric MDI-QKD based on the satellite in the case of statistical fluctuations, and uses the local search algorithm to achieve full parameter optimization under the condition of considering the probability of sending the signal. Compared with fixed related parameters, the key rate is increased by an order of magnitude. On this basis, random forest is used to predict the high-precision optimal parameters, thereby eliminating the simulation and iteration required by the search method to meet the real-time optimization of the future QKD network.
The mode-pairing quantum key distribution (MP-QKD) achieves high key generation rates and high security over long-distance, low-loss optical fibers through its mode-pairing technique. In practical applications, when considering finite data sets, MP-QKD requires optimization of parameters such as channel optical bit error rate ( e opt ), detector’s dark count( P d ), the total number of signals( N ), and the number of phase slices( M ). However, traditional exhaustive traversal or local search algorithms do not meet the time and resource requirements of real-time communication systems. Combined with machine learning, the quantum key distribution for parameter optimization prediction has become the mainstream approach. XGBoost, a gradient boosting decision tree (GBDT) algorithm, random forest (RF) representing a classic bagging-based ensemble learning method, and back-propagation neural networks (BPNN) are important algorithms in neural networks. This study employs the LightGBM (light gradient boosting machine) algorithm to predict optimized parameters for MP-QKD. The results demonstrate that LightGBM efficiently and accurately predicts optimal parameters, slightly outperforming XGBoost, RF, and BPNN in parameter prediction, and can serve as a reference for future real-time QKD networks.
Abstract As the number of users in quantum key distribution (QKD) networks continues to grow, parameter selection (e.g., probability of X-bases, signal state strength, and decoy state levels) has become increasingly complex. Traditional methods such as local search algorithms (LSA) are commonly employed for parameter optimization but suffer from significant computational overhead, slow execution on low-power platforms, and resultant system delays. These limitations hinder the scalability and real-time requirements of large-scale QKD networks. To address these challenges, this paper proposes an innovative approach that combines Bayesian optimization with XGBoost model to optimize QKD system parameters. The proposed model is evaluated against random forest (RF) and back-propagation neural network (BPNN) models based on three critical metrics: prediction accuracy, model fitting, and training efficiency. Simulation results demonstrate that the proposed model significantly outperforms both RF and BPNN in terms of accuracy and computational efficiency. Notably, while achieving predictive performance comparable to LSA, the proposed method reduces computation time by three orders of magnitude. This research introduces a novel and efficient framework for QKD parameter optimization, providing new insights and directions for enhancing the performance, scalability, and real-time adaptability of QKD systems.
The parameter configuration of quantum key distribution (QKD) has a great effect on the communication effect, and in the practical application of the QKD network in the future, it is necessary to quickly realize the parameter configuration optimization of the asymmetric channel measurement-device-independent QKD according to the communication state, so as to ensure the good communication effect of the mobile users, which is an inevitable requirement for real-time quantum communication. Aiming at the problem that the traditional QKD parameter optimization configuration scheme cannot guarantee real-time, in this paper we propose to apply the supervised machine learning algorithm to the QKD parameter optimization configuration, and predict the optimal parameters of TF-QKD and MDI-QKD under different conditions through the machine learning model. First, we delineate the range of system parameters and evenly spaced (linear or logarithmic) values through experimental experience, and then use the traditional local search algorithm (LSA) to obtain the optimal parameters and take them as the optimal parameters in this work. Finally, we train various machine learning models based on the above data and compare their performances. We compare the supervised regression learning models such as neural network, K-nearest neighbors, random forest, gradient tree boosting and classification and regression tree (CART), and the results show that the CART decision tree model has the best performance in the regression evaluation index, and the average value of the key rate (of the prediction parameters) and the optimal key rate ratio is about 0.995, which can meet the communication needs in the actual environment. At the same time, the CART decision tree model shows good environmental robustness in the residual analysis of asymmetric QKD protocol. In addition, compared with the traditional scheme, the new scheme based on CART decision tree greatly improves the real-time performance of computing, shortening the single prediction time of the optimal parameters of different environments to the microsecond level, which well meets the real-time communication needs of the communicator in the movable state. This work mainly focuses on the parameter optimization of discrete variable QKD (DV-QKD). In recent years, the continuous variable QKD (CV-QKD) has developed also rapidly. At the end of the paper, we briefly introduce academic attempts of applying machine learning to the parameter optimization of CV-QKD system, and discuss the applicability of the scheme in CV-QKD system.
Measurement-device-independent quantum key distribution (MDI-QKD) has been demonstrated in both laboratories and field tests using attenuated lasers combined with the decoy-state technique. Although researchers have studied various decoy-state MDI-QKD protocols with two or three decoy states, a clear comparison between these protocols is still missing. This invokes the question of how many types of decoy states are needed for practical MDI-QKD. Moreover, the system parameters to implement decoy-state MDI-QKD are only partially optimized in all previous works, which casts doubt on the actual performance of former demonstrations. Here, we present analytical and numerical decoy-state methods with one, two, and three decoy states. We provide a clear comparison among these methods and find that two decoy states already enable a near-optimal estimation and more decoy states cannot improve the key rate much in either asymptotic or finite-data settings. Furthermore, we perform a full optimization of system parameters and show that full optimization can significantly improve the key rate in the finite-data setting. By simulating a real experiment, we find that full optimization can increase the key rate by more than one order of magnitude compared to nonoptimization. A local search method to optimize efficiently the system parameters is proposed. This method can be four orders of magnitude faster than a trivial exhaustive search to achieve a similar optimal key rate. We expect that this local search method could be valuable for general fields in physics.
For a practical quantum key distribution (QKD) system, parameter optimization, the choice of intensities and the probabilities of sending them, is a crucial step in gaining optimal performance, especially when one realistically considers a finite communication time. With the increasing interest in the field to implement QKD over free space on moving platforms, such as drones, handheld systems, and even satellites, one needs to perform parameter optimization with low latency and with very limited computing power. Moreover, with the advent of the internet of things, a highly attractive direction of QKD could be a quantum network with multiple devices and numerous connections, which provides a huge computational challenge for the controller that optimizes parameters for a large-scale network. Traditionally, such an optimization relies on brute-force search or local search algorithms, which are computationally intensive, and will be slow on low-power platforms (which increases latency in the system) or infeasible for even moderately large networks. In this work we present a method that uses a neural network to directly predict the optimal parameters for QKD systems. We test our machine learning algorithm on hardware devices including a Raspberry Pi 3 single-board computer (similar devices are commonly used on drones) and a mobile phone, both of which have a power consumption of less than 5 W, and we find a speedup of up to two to four orders of magnitude when compared to standard local search algorithms. The predicted parameters are highly accurate and can preserve, e.g., over 95%--99% of the optimal secure key rate for a given protocol. Moreover, our approach is highly general and can be applied effectively to various kinds of common QKD protocols.