Description
Date depot: 23 septembre 2022
Titre: AI-based Optical wireless communication device fingerprint in future network
Directeur de thèse:
Xun ZHANG (LISITE)
Domaine scientifique: Sciences pour l'ingénieur
Thématique CNRS : Signal et communications
Resumé: Under the support of I H2020 European project 6G BRAINS (https://5G-ppp.eu/6G-brains/), ISEP proposes a PhD project on “AI-based Optical wireless communication device fingerprint extraction method in the future network”.
5G wireless networks are the next generation of telecommunication systems, they have gained momentum to connect billions of people as well as things, whether on motion or attached to an infrastructure. 5G networks are expected to support massive user connections and exponentially increase wireless services. In addition, due to a tremendous number of Internet-of-Things (IoT) devices featured by the massive Machine-Type Communication (mMTC) extensive applications in the 5G network, high data rates, high connection density and ultra-reliable low latency communication (URLLC) should be provided. Meanwhile, Cybersecurity has become one of the major concerns in all types of systems, from data centers to edge computing and Internet of Things. In this broad context, the proposed research work focuses on the confidentiality of embedded wireless systems, i.e. on preventing the leakage of data that could be intercepted by an attacker. Hence, a circuit’s architecture, specifically its low-level implementation, is ultimately responsible of its security, as it plays a key linking role mapping inputs to outputs of a wireless communication system.
Device fingerprinting has emerged as a potential solution to reduce the vulnerability of rogue devices and node forgery. Device fingerprinting is the process of gathering device information to generate device-specific signatures and using them to identify individual devices. Its low-complexity and difficult or impossible to forge property could be perfectly matched with the security requirements of the 5G network. Notably, wireless device identification via Radio Frequency (RF) fingerprint becomes a widely concerned physical-layer security mechanism and has been investigated in Wi-Fi, LTE (Long Term Evolution) and ZigBee systems. It provides an opportunity to accomplish the authentication and the target device identification in the physical layer [Burchardt 2014].
Inspired by the success of RF fingerprinting technology, device fingerprinting technology has been carried out in the field of VLC communication to improve the security of VLC communication system. VLC relies on the visible light (VL) spectrum for communication rather than the RF spectrum. VL is not regulated which can be used freely and therefore significantly reducing the cost for providers [Xu2011].
ISEP has proposed a study [Shi2020] of a novel device fingerprint extraction and identification method to improve the security of VLC in 5G networks as shown in Figure.1. This method extracts the unique fingerprints of LEDs from their emitted signals. The fingerprints of four identical white LEDs were extracted successfully from the received 5G NR (New Radio) signals. To perform identification four types of machine learning-based classifiers were employed and the resulting accuracy was up to 97.1\%.
The objectif of this PhD thesis is do research on the study of device fingerprint for VLC system in 5G and beyond 5G. In this context, this subject address three main objectives:
Optical wireless communication device fingerprint state-of-the-art
AI-based OWC device fingerprint extraction method definition and experimentation setup
Evaluation of proposed method in IoT network
Doctorant.e: Liu Ziqi