Non-orthogonal multiple access (NOMA) has been recognized as a promising multiple access technique for enhanced spectral efficiency in the current and next-generation wireless networks. In this paper, we examine a realistic NOMA model where users, assisted by a regenerative relay, cannot be fully trusted. We address the challenge of ensuring secure access for these users while accounting for the error propagation in successive interference cancellation (SIC) during the decoding process. For such, we formulate and solve two optimization problems, viz. maximizing the minimum secrecy rate of the users and maximizing the sum secrecy rate of the users, while accounting for SIC errors and the constraint on the power budget. For each case, we derive the optimal power allocation solution to achieve positive secrecy rates despite imperfect SIC. Simulation results provide key insights on the obtained secrecy rates and power allocations, factoring in residual interference. The joint optimal solution for the decoding order and power allocation is compared with different benchmark schemes: optimal decoding order and equal power allocation, fixed decoding order and equal power allocation, fixed decoding order and optimal power allocation, and optimal decoding order and channel-based power allocation. Our proposed framework demonstrates average performance gains of about 47.62 dB, 50.79 dB, 54.02 dB and 39.83 dB over these schemes and, hence, the fact that the proposed framework can substantially improve the secrecy performance.

中文翻译：

---

不完美 SIC 下行链路协作 NOMA 中不可信接收机的功率分配和译码顺序选择保密公平性


非正交多址 （NOMA） 已被公认为一种很有前途的多址技术，可以提高当前和下一代无线网络的频谱效率。在本文中，我们研究了一个现实的 NOMA 模型，在该模型中，在再生继电器的协助下，用户不能完全信任。我们解决了确保这些用户安全访问的挑战，同时考虑了解码过程中连续干扰消除 （SIC） 中的错误传播。为此，我们制定并解决了两个优化问题，即最大化用户的最小保密率和最大化用户的总保密率，同时考虑 SIC 错误和对功率预算的约束。对于每种情况，我们都推导出了最佳功率分配解决方案，以便在 SIC 不完美的情况下实现正保密率。仿真结果提供了有关所获得的保密率和功率分配的关键见解，其中考虑了残余干扰。将译码顺序和功率分配的联合最优解与不同的基准方案进行了比较：最优译码顺序和相等功率分配、固定译码顺序和相等功率分配、固定解码顺序和最优功率分配、最优解码顺序和基于通道的功率分配。我们提出的框架表明，与这些方案相比，平均性能增益约为 47.62 dB、50.79 dB、54.02 dB 和 39.83 dB，因此，所提出的框架可以显著提高保密性能。