Optimal Parameter Prediction for Secure Quantum Key Distribution Using Quantum Machine Learning Models

Abstract

The advent of quantum computing is bringing threats to successful operations of classical cryptographic techniques. To conduct quantum key distribution (QKD) in a finite time interval, there is a need to estimate photon states and analyze the fluctuations statistically. The use of brute force and local search methods for parameter optimization are computationally intensive and becomes an infeasible solution even for smaller connections. Therefore, the use of quantum machine learning models with self-learning ability is useful in predicting the optimal parameters for quantum key distribution. This chapter discusses some of the quantum machine learning models with their architecture, advantages, and disadvantages. The performance of quantum convoluted neural network (QCNN) and Quantum Particle Swarm Optimization (QPSO) towards QKD is found to be good compared to all the other quantum machine learning models discussed.