MISO Secure Transmission with Imperfect Channel State Information

In this paper, we study artificial noise (AN) assisted beamforming secure transmission system with imperfect main channel state information (CSI) between a friendly cooperative jammer (Oscar) and an authorized receiver (Bob). We deduce out the maximum secrecy rate of secure system and the corresponding optimal power allocation ratio between information signal and AN under the constraint of secrecy outage probability. For realistic security communication system, we use advanced back propagation neural network (BPNN) for channel estimation. The numerical results show that estimation error results in a decrease in secrecy rate, but BPNN channel estimator still can guarantee secure transmission. Analytical derivations and numerical simulations are presented to validate the correctness of obtained expressions.