Secrecy Capacity Analysis Over \alpha - \mu Fading Channels

Abstract

In this letter, we study the secrecy capacity of the classic Wyner's model over the &x03B1; - &x03BC; fading channels, where &x03B1; and &x03BC; specify the nonlinearity and clustering of fading channels, respectively. The average secrecy capacity (ASC) is derived in closed-form by using the extended generalized bivariate Fox's H-function. Moreover, the asymptotic analysis of ASC in high signal-to-noise ratio (SNR) regime is conducted. The asymptotic results unveil that the ASC follows the scaling law of &x0398; (ln &x03C1;), where p stands for the ratio between the average powers of main channels and eavesdropping channels. Moreover, the ASC can be enhanced by increasing the transmit SNR, while there exists a ceiling of ASC as the SNRs at both sides are improved simultaneously. The accuracy of the analytical results is validated by Monte-Carlo simulations. The numerical results show that rigorous fading channels are beneficial to the secrecy performance, that is, serious nonlinearity (small &x03B1;) and sparse clustering (small &x03BC;) will lead to the improvement of ASC.