Two-way Training Scheme For Discriminatory Channel Estimation In Physical Layer Security

Eavesdropping has become a potential security threat in the wireless communication due to the increasingly developed information techniques,especially as the popular usage of the internet and wireless networks.Traditional encoding and decoding methods cannot guarantee the perfect security of wireless communications once the password is cracked by an illegal attacker,which surely would damage the benefit of the legal user.As an alternative technique,the emerging security method from the physical layer tries to ensure a perfect security via exploiting the unique feature of the legitimate channel information.Physical layer security methods have a solid foundation based on the information theory to realize perfect security communication.Hence,the physical layer security has attracted more and more attentions.Following are the contributions of this paper:1)We develop a two-way training scheme for Discriminatory Channel Estimation(DCE)under the MIMO wireless networks and the MIMO-relay co-operative networks,respectively.The proposed scheme includes two phases: reverse training and forward training,which is to ensure the security of channel estimation process between the transmitter and the legitimate receiver(LR).Specifically,it makes use of properly designed training sequence to ensure the channel estimation at the LR which in turn limits the Un-authorized Receiver's(UR)channel estimation during the training data transmission.2)In this paper,we propose a new two-way training scheme for DCE based on the Linear Minimum Mean-Square Error(LMMSE)method.To characterize the performance of DCE,a closed-form expression of the Normalized Mean Squared Error(NMSE)on the channel estimation is derived both for the LR and the UR,respectively.Furthermore,the developed analytical results based on NMSE are utilized to perform optimal power allocation between the training pilots and the Artificial-Noises(AN)under the MIMO systems and MIMO-Relay systems,respectively.