Secure MIMO D2D communication based on a lightweight and robust PLS cipher scheme

Abstract

The Multiple-Input Multiple-Output (MIMO) technology is one of the key enablers of the 5G system. This technology allows the concurrent transmission of data using multiple antennas, which increases the channel robustness, overall throughput and link capacity. The security of 5G systems is very critical and it has received serious attention from various research groups. To that end, physical layer security (PLS) has recently emerged as a promising candidate to enhance network security, and several PLS schemes have been proposed for MIMO systems. Most of these MIMO PLS solutions target data confidentiality, however, they suffer from a main limitation due to relying solely on the random parameters of the channel. To overcome this limitation, we propose a PLS cipher solution for D2D MIMO systems, which is based on the dynamic key structure and several other factors including a private key and a physical channel parameter. Both of these parameters are combined to produce a dynamic key, which is used to generate the cipher primitives that are used to encrypt or decrypt data frames. For every input frame, the cipher primitives are updated, and for each frame symbol, the selection cipher primitives are updated. The update process is designed to enhance the security level and to ensure the avalanche effect in a simple and lightweight manner. Consequently, the proposed technique achieves high immunity against existing attacks due to the dynamic property which complicates the attacker’s tasks. A high efficiency level is also attained since a single round with simple cipher operations are employed. The proposed scheme decreases the required latency, resources and overhead in comparison to security schemes at upper layers of the D2D MIMO communication system. Finally, we conducted several security and performance tests to evaluate the proposed solution and to prove its robustness and efficiency. © 2020, Springer Science+Business Media, LLC, part of Springer Nature.