Sualtı kuantum anahtar dağıtım sistemlerinin derin öğrenme tabanlı optimizasyonu

The rapid advancements on quantum computers with high computational capabilities have raised concerns regarding the use of classical cryptography methods. As we enter the era of advanced quantum computing, traditional key generation methods commonly used in wireless systems will encounter substantial security vulnerabilities. These methods rely on computational complexity assumptions, which render them susceptible to attacks by powerful computers. In contrast, Quantum Key Distribution (QKD) technique leverages the principles of quantum mechanics, enabling it to provide a high level of security that is theoretically unconditional. One of the emerging applications of QKD is to provide quantum-secure communication in maritime missions, such as submarine-to-submarine communication, autonomous underwater vehicle (AUV) data offloading, underwater sensor network (USN) data fusion, etc. In this study, we consider underwater QKD systems with time-gated single photon avalanche photodiode (SPAD) and present a comprehensive performance analysis and optimization. The main contribution of this research study is to investigate the quantum bit error rate (QBER) performance of the well-known QKD protocol, namely the BB84 protocol, with respect to different system and transceiver parameters in underwater channels. Our aim is to optimize the bit time and field of view (FoV) parameters in order to minimize the QBER performance metric. Through a meticulous analysis of the propagation delay and angle of arrival results, we determine the bit time and FoV, respectively, by striking a balance between the average number of received photons and background noise. Furthermore, our study provides critical insights into determining the optimal gate time in an underwater QKD system based on the optimal bit time and FoV to minimize the QBER. Given the inherent computational complexity associated with this optimization process, we identify the most influential transceivers and channel parameters that exert an impact on the determination of bit time and FoV, and utilize a deep learning model for the system optimization. We first perform Monte Carlo simulations for a subset of possible scenarios and train the deep learning model on them. Then, we use this model to extract the optimum values for possible underwater scenarios.