Two-way discriminatory channel estimation for non-reciprocal wireless MIMO channels

Abstract

The idea of discriminatory channel estimation (DCE) was recently proposed to differentiate the performance at a legitimate receiver (LR) and an unauthorized receiver (UR) by carefully designing the training phase of transmission. In particular, an artificial noise (AN) signal is inserted into the training signal to degrade the channel estimation performance at UR. However, without careful design, the inserted AN may potentially leak into LR's channel causing degradation in LR's performance. In order to minimize such an undesirable AN leakage, the transmitter must acquire sufficient knowledge of the downlink channel to LR for careful signalling, which is particularly difficult to achieve in non-reciprocal channel environments. In this work, we design a DCE scheme with multiple training stages to achieve this task. The key feature of the proposed scheme is to utilize a two-way training scheme that allows both the transmitter and LR to collaboratively send different training signals to achieve DCE. Simulation results demonstrate that our proposed scheme is able to achieve accurate channel estimation and data detection performances at LR, while severely degrading UR's performances.