I. Introduction
Security and spectral efficiency are two key issues that restrict the development of future wireless communications, and they have become research hotspots for some time [1]–[5]. Traditionally, security issues have been addressed in the upper layers of the network protocol stack, like the data encryption standard (DES) and the advanced encryption standard (AES), which are based on the mathematical computational hardness. With the substantial increase in computing power and the development of quantum computing in the near future, conventional encryption techniques will become increasingly unreliable, let alone the difficulties and the security vulnerabilities in secret key distribution and management [6]. As an alternative or a complement to the conventional security technologies, the physical-layer security has been received tremendous attentions since the notion of secrecy capacity [7] was put forward, by exploiting characteristics of the channels (e.g., fading, noise, and interference). More importantly, the physical-layer security has been proven to be reliable from an information-theoretic perspective [8].