Machine Learning Techniques for Detecting Attackers During Quantum Key Distribution in IoT Networks With Application to Railway Scenarios

Internet of Things (IoT) deployments face significant security challenges due to the limited energy and computational power of IoT devices. These challenges are more serious in the quantum communications era, where certain attackers might have quantum computing capabilities, which renders IoT devices more vulnerable. This paper addresses the problem of IoT security by investigating quantum key distribution (QKD) in beyond 5G networks. An algorithm for detecting an attacker between a transmitter and receiver is proposed, with the side effect of interrupting the QKD process while detecting the attacker. Afterwards, Artificial neural network (ANN) and deep learning (DL) techniques are proposed in order to detect the presence of an attacker during QKD without the need to disrupt the key distribution process. An architecture for implementing QKD in beyond 5G IoT networks is proposed, offloading the heavy computational tasks to IoT controllers. In addition, an implementation scenario for securing IoT communications for sensors deployed in railroad networks is described. The results show that the proposed ML techniques can reach 99% accuracy in detecting attackers.