Physical Layer Authentication in Private Campus Networks based on Machine Learning

Mobile communication is in the deployment phase of its fifth generation (5G) technology standards and the research activities have begun toward the sixth generation (6G) of mobile communication systems. Private campus networks (PCNs) serving a specific use case of the vertical industries (e.g., smart factory) are commonplace in the 5G and will continue to be the same in the future. Facilitating secure operations in such a PCN is of the utmost priority. Physical layer security (PLS) is one of the most promising security paradigms for the present and future generations of mobile communication systems, which works on the principle of channel reciprocity. In this paper, PLS concepts are applied to design authentication schemes for the PCN. Measurements of the physical layer are carried out in a live PCN and a grid-wise radio environment map is constructed. A lookup table-based as well as Machine learning (Decision trees, Random Forest) based PHY packet authentication schemes are designed and evaluated to showcase their applicability in providing efficient authentication in the PCN environment.