Abstract

Artificial noise (AN) is an efficient way to enhance physical layer security of multiantenna wiretap channels with a passive eavesdropper (ED) whose channel state information is unavailable. However, the AN is known to be vulnerable to the eavesdropping attack with a large antenna array. To tackle the problem, the artificial fast fading (AFF) scheme has been developed to deliberately randomize the ED's effective channel so as to disrupt its channel estimation. Some recent progress has been made for the secrecy-rate performance of the AFF scheme in the multi-input single-output wiretap channels. However, it turns out that the performance is over-estimated. In this correspondence, we provide a comprehensive analysis on the secrecy-rate of the AFF scheme in general multi-input multi-output wiretap channels. Unfortunately, our analysis reveals a negative result that the conventional AFF is no better than the AN methods, without some pre-shared information among the legitimate parties. This is in fact in stark contrast to the previous observations. Motivated by our analysis, we propose a novel cross layer design by combining the AFF in the physical layer with the secret-keys in the upper layer crypto systems. Then, we mathematically prove that our method asymptotically nullifies the information leakage to the ED in the physical layer for any number of ED antennas even without the aid of the wiretap codes, while further enhancing the upper layer security, and thus provides an enhanced security over the layers. Finally, numerical results demonstrate the accuracy of our analysis and the efficiency of the proposed scheme.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call