A semiblind discriminatory channel estimation training scheme in MIMO two-way full duplex relay networks

Discriminatory channel estimation (DCE) is a recently developed scheme to enlarge the performance difference between legitimate terminal and unauthorized terminal. Specifically, it makes use of properly designed training signals to maintain channel estimation performance at legitimate terminal while limiting the unlawful terminal´s eavesdropping capability during data transmission. In this paper, we extend our previous work, namely, a semiblind whitening-rotation (WR) based two-way training scheme into a MIMO two-way duplex-relay networks. For this, an artificial noise strategy is employed to jam the eavesdropper which in return keep a minimal level of interference to the legitimate terminal in the systems. Furthermore, a power allocation solution between the training signal and artificial noise (AN) is offered to optimal DCE performance. Results demonstrate that the two-way DCE scheme can be successfully extended into the MIMO relay systems without any priori knowledge of system channel statement information.