Precoding Adjustment Based Physical Layer Secret Key Generation for MU-MIMO Systems

Physical layer key generation in multi-user multiple-input multiple-output (MU-MIMO) system is challenging due to the poor randomness of the equivalent channel response. The precoding matrix combines the channels from different antennas in a constructive way, resulting in a flat fading equivalent channel. In this paper, we proposes new physical layer key generation schemes based on precoding adjustment. Two strategies are proposed to enhance the randomness of the equivalent channel response at the cost of beamforming gain, which consist in adjusting random elements and adjusting the largest elements. When considering the transmission of artificial noise (AN) at the base station, random phase adjustment strategy is proposed to the precoding scheme which does not affect the beamforming gain. The signal-to-interference-plus-noise ratio of user and key generation performance are simulated. Simulation results show that adjusting the precoding elements with larger amplitudes is more efficient to increase the information entropy of the channel response. When the system is able to transmit an AN with high power ratio, random phase adjustment can help to generate secret keys with higher information entropy compared to two other strategies.