Channel-Adaptive Privacy Enhancement for NOMA-Based Antagonistic Overlay Cognitive Networks

Overlay cognitive networks based on non-orthogonal multiple access (NOMA) can introduce substantial privacy concerns, especially in antagonistic systems where primary and secondary networks lack mutual trust. This paper highlights two critical privacy challenges and investigates a NOMA-assisted purely antagonistic overlay cognitive network. As part of our privacy design, we propose a Channel-Adaptive Dual-Phase Cooperative Jamming (CADP-CJ) strategy, leveraging reverse successive interference cancellation and a dynamic top-down power allocation approach based on the available channel-state information. The ergodic secrecy rate (ESR) for both single-user and multi-user scenarios is derived in closed form by means of Taylor-McLaurin expansions and Gaussian-Chebyshev quadrature, while considering Nakagami-m fading across all channels. Furthermore, the closed-form expressions for the asymptotic ESR are presented to provide deeper insights. The accuracy of our analytical results is corroborated through Monte-Carlo simulations, which also confirm that our scheme ensures a positive ESR in both single and multi-user cases. We comprehensively analyze the impact of the fading properties of the channels involved and comment on optimal jamming power using the CADP-CJ strategy. Notably, our proposed system outperforms benchmark systems, particularly those based on orthogonal multiple access, with an 86% enhancement for primary users and 64% for secondary users.