The eigenvalue decomposition method for wireless physical layer security

For the wiretapping in wireless communication system, this paper proposes an eigenvalue decomposition method for wireless physical layer security. In the method, firstly, the channel state information matrix is decomposed into sub-channels by eigenvalue decomposition. Secondly, we choose the sub-channels with bigger eigenvalue to transmit useful signal and the one with the smallest eigenvalue is transformed into a random channel. Then, the received signal of legitimate receiver becomes steady, while eavesdropper´s turns random. The simulations show that, by choosing the characteristic, the method can decrease the demands of legitimate channel state, and keep the eavesdropper´s error probability at a high level, so as to increase the system security. We can get the conclusion that the eigenvalue decomposition method can prevent wiretapping and increase wireless physical layer security between the transmitter and legitimate receiver effectively.