Artificial Noise-Aided User Scheduling for Optimal Secrecy Multiuser Diversity

Abstract

In this letter, we propose a user scheduling scheme, which enables non-scheduled users to generate artificial noise in order to enhance wireless security in a multiuser wiretap network, where there exist $N$ users, a single desired receiver, and $K$ eavesdroppers. Our proposed scheme can achieve the optimal multiuser diversity gain (i.e., $\log \log N$ ) by adopting appropriate threshold values for scheduling feedback, which is also able to significantly reduce signaling overhead. We mathematically prove the achievability based on the derivation of a secure user-scaling law . Simulation results demonstrate that the proposed scheme achieves a better average secrecy rate than the conventional scheme in a finite $N$ case, even if both schemes obtain the same secure user-scaling law as $N \rightarrow \infty $ .