[Paper Review] UAV Secure Downlink NOMA Transmissions: A Secure Users Oriented Perspective

Hui‐Ming Wang, Xu Zhang|arXiv (Cornell University)|Jun 9, 2020

Advanced Wireless Communication Technologies4 citations

TL;DR

This paper proposes a UAV-aided downlink NOMA system that maximizes the minimum secrecy rate among security-required users (SUs) by jointly optimizing user scheduling, power allocation, and UAV trajectory, while ensuring QoS for potential eavesdropping QoS-required users (QUs). The scheme significantly outperforms conventional OMA in secrecy rate, especially under dynamic UAV mobility and constrained resources.

ABSTRACT

This paper proposes a secure downlink multi-user transmission scheme enabled by a flexible unmanned aerial vehicle base station (UAV-BS) and non-orthogonal multiple access (NOMA). According to their heterogeneous service requirements, multiple legitimate users are categorized as security-required users (SUs) and quality of service (QoS)-required users (QUs), while these QUs can potentially act as internal eavesdroppers which are curious about the secrecy transmissions of SUs. In such a context, our goal is to maximize the achievable minimum secrecy rate among SUs through the joint optimization of user scheduling, power allocation, and trajectory design, subject to the QoS requirements of QUs and the mobility constraint of UAV-BS. Due to the non-convexity of the problem, an efficient iterative algorithm is firstly proposed, based on the alternative optimization (AO) and successive convex approximation (SCA) methods and along with a penalty-based algorithm to deal with the introduced binary integer variables, to obtain a sub-optimal solution. Then, we propose an SUs-oriented low-complexity algorithm by taking advantage of the inherent characteristics of the optimization problem, which can efficiently reduce the computational complexity and can act as a reasonable initial solution for the previous iterative algorithm to achieve better performance. Finally, the superiority of our proposed scheme compared with the conventional orthogonal multiple access (OMA) one is validated by numerical simulation results.

Motivation & Objective

• Address the challenge of physical-layer security in UAV-aided downlink NOMA systems where QoS-required users (QUs) may act as internal eavesdroppers.
• Maximize the minimum secrecy rate among security-required users (SUs) to ensure fairness and robustness in secrecy performance.
• Account for QoS constraints of QUs and mobility limitations of the UAV-BS in the optimization framework.
• Develop low-complexity and iterative algorithms to balance performance and computational efficiency in non-convex optimization.
• Demonstrate the superiority of NOMA over OMA in secrecy rate and resilience to internal eavesdropping threats.

Proposed method

• Formulate a non-convex optimization problem combining user scheduling, power allocation, and UAV trajectory design under QoS and mobility constraints.
• Apply alternative optimization (AO) and successive convex approximation (SCA) to iteratively solve the non-convex problem by approximating non-convex terms.
• Introduce a penalty-based method to handle binary integer variables introduced by user scheduling decisions.
• Propose a low-complexity SUs-oriented algorithm leveraging problem structure to reduce computational load and serve as an initial solution for the iterative algorithm.
• Use successive interference cancellation (SIC) at receivers to decode superimposed NOMA signals based on channel gain differences.
• Model the A2G LoS channel to exploit UAV mobility for dynamic channel gain control and improved spectral efficiency.

Experimental results

Research questions

• RQ1 How can UAV-BS trajectory and power allocation be jointly optimized to maximize secrecy rate among SUs in a NOMA downlink system?
• RQ2 What is the impact of QoS-required users (QUs) acting as internal eavesdroppers on the secrecy performance of SUs?
• RQ3 How do system parameters such as flight period, transmit power, and QoS requirements affect the secrecy rate and UAV trajectory design?
• RQ4 Can a low-complexity algorithm achieve near-optimal secrecy performance while reducing computational cost?
• RQ5 To what extent does NOMA outperform OMA in securing transmissions when internal eavesdroppers are present?

Key findings

• The proposed iterative algorithm based on AO and SCA achieves a sub-optimal solution for the non-convex secrecy rate maximization problem.
• The low-complexity SUs-oriented algorithm reduces computational complexity significantly and provides a high-quality initial solution for the iterative algorithm.
• Numerical results show that the proposed NOMA scheme achieves significantly higher secrecy rates than conventional OMA, especially under dynamic UAV mobility.
• As flight period increases, secrecy performance converges to a stationary value due to UAV-BS hovering at optimal locations, indicating diminishing returns beyond a certain duration.
• Higher available transmit power allows UAV-BS to fly further from QUs, degrading their eavesdropping capability and improving SU secrecy rates.
• Stricter QoS requirements for QUs force the UAV-BS to move closer to them, which degrades SU channel quality and reduces secrecy performance.