Exploiting Artificial Randomness for Fast Secret Key Generation in Quasi-static Environments

Abstract

This paper investigates a secret key generation problem in quasi-static wireless communication environments. To increase the key generation rate, we propose an efficient artificial randomness (AR) aided key generation approach in time division duplexing (TDD) multiple-input single-output multiple-eavesdropper (MISOME) communication systems, adopting two transmitter precoding techniques including zero-forcing (ZF) and maximum ratio transmission (MRT). We prove that the proposed approach outperforms existing key generation approaches without AR in terms of increasing achievable key rate. Furthermore, the improved key rate is derived to be bounded by the entropy of legitimate users' and eavesdroppers' channels in MRT precoded systems, while the improved key rate increases linearly with the number of channel probing rounds in ZF precoding systems. Our proposed AR-aided approach needs no extra nodes to assist in key generation and can resist eavesdroppers with multiple antennas. Numerical results show that with the increase of AR variation range, ZF obtains a higher key generation rate while the key rate of MRT tends to a lower fixed value.