Physical layer security for beyond 5G/6G networks: Emerging technologies and future directions
Fauzia Irram,Mudassar Ali,Muhammad Naeem,Shahid Mumtaz
DOI: https://doi.org/10.1016/j.jnca.2022.103431
IF: 7.574
2022-10-01
Journal of Network and Computer Applications
Abstract:Physical layer security (PLS) has proven to be a potential solution for enhancing the security performance of future 5G networks, which promises to fulfill the demands of increasing user traffic. Preventing eavesdroppers from overhearing and stealing useful information in such high traffic environments is as challenging as eliminating them from the network. The goal of this survey is to present a comprehensive study of the latest PLS works proposed to enhance the security performance in different 5G technologies. The survey starts by first giving a detailed introduction and overview of existing surveys that explicitly or partially discuss PLS in 5G and its emerging technologies. Many researchers have presented a number of PLS schemes, using either a separate technology such as Multiple-input–multiple-output (MIMO), Millimeter Wave (mmWave), Radio frequency (RF), Non-orthogonal multiple access (NOMA), Visible light communication (VLC), etc., or a combination of two or more technologies, for securing each field of future 5G networks such as Heterogeneous networks (HetNets), Device-to-Device (D2D), Internet-of-Things (IoT), Cognitive radio network (CRN), Unmanned Aerial Network (UAV), etc. After summarizing the existing surveys, we present a detailed overview on the PLS research works performed till now in HetNets, with respect to its different underlaying technologies, as well as in other emerging 5G technologies. Then, optimization ontology is presented that discusses different security metrics used for measuring PLS performance. Different from rest of the surveys, our survey includes a comprehensive discussion regarding the proposed PLS techniques based on artificial intelligence and machine learning techniques, especially highlighting the works performed using reinforcement learning and deep learning algorithms, allowing us to understand how artificial intelligence can help to achieve better PLS. Towards the end, we discuss numerous challenges being encountered in practical implementation of PLS techniques, and propose different interesting areas that can be opted as future research direction.
computer science, interdisciplinary applications, software engineering, hardware & architecture