An Artificial-Noise-Aided Anti-Eavedropping Design for OFDM Systems With Imperfect Channel State Information

This article investigates an anti-eavesdropping scheme named orthogonal frequency-division multiplexing with artificial noise (OFDM-AN) based on orthogonal frequency-division multiplexing (OFDM) and artificial noise (AN). In the proposed scheme, all the subcarries are divided into many different groups with a fixed carrier number and the AN is generated by a differential method in each group. First, the transmitter and receiver of OFDM-AN are introduced. Then, both the secrecy rate (SR) and bit error rate (BER) of OFDM-AN with imperfect channel state information (CSI) are introduced. Simulation results show that the more accuracy the channel estimation is, the better the system performance will be. Finally, the power allocation between information signals and AN of the proposed scheme is analyzed. Simulation results show that if information signals and AN are allocated equal power, the scheme obtains the maximum SR. Furthermore, simulation curves and theoretical curves of the BER and SR match very well, respectively. Theoretical analysis and simulation results show that the proposed scheme can provide security for OFDM-based systems even with imperfect CSI.