NOMA-Enabled Underlay Cognitive IoT Networks: Secrecy Energy Efficiency Optimization and Deep Learning-Based Assessment

Abstract

Physical layer security (PLS) is a novel approach that has surfaced as an additional security measure for wireless networks, enhancing the existing cryptography-based techniques. The notion of secrecy energy efficiency (SEE) effectively addresses the task of establishing secure and energy-efficient communication in wireless networks. Nonorthogonal multiple access (NOMA)-enabled cognitive Internet-of-Things (IoT) systems have been proposed to enhance spectrum utilization efficiency, connectivity, and quality of service for IoT applications, where IoT devices can function as secondary users (SUs) and take advantage of the spectrum used by primary users (PUs) for communication among themselves or with the Internet. The objective of this paper is to study the SEE performance of NOMA-enabled underlay cognitive radio networks (NOMA-UCRNs) with an external passive eavesdropper. Initially, analytical models are developed to assess the SEE and secrecy sum rate (SSR) of the secondary network in NOMA-UCRN, considering imperfect successive interference cancellation conditions and interference power constraints of the primary receiver. Next, we determine the optimal transmit power allocation (OTPA) for the SUs at the secondary base station that maximizes the SEE while adhering to the requirements of maintaining minimum data rates for the SUs and satisfying the interference constraint on the primary receiver. The OTPA is determined by employing an iterative algorithm based on the Dinkelbach method. It is demonstrated that the proposed OTPA strategy enhances the following: (i) the SEE of the secondary network by 190% and 360% and (ii) the secrecy sum rate by 30% and 95% compared to random transmit power allocation and equal transmit power allocation strategies, respectively. Lastly, a state-of-the-art deep neural network (DNN) framework is devised to achieve precise and rapid prediction of the OTPA, ultimately maximizing the SEE of the NOMA-UCRN under investigation.