This paper studies enhancing non-orthogonal multiple access (NOMA) spectral efficiency (SE) through optimizing the assigned power to each NOMA user depending on their channel states. This would improves the overall NOMA system performance in an adaptive manner in accordance with the channel state variations. An optimization problem for SE maximization is presented to optimize the performance of NOMA system. To solve the formulated problem, three power assignment schemes are proposed; firstly, a constant weight based water- filling approach is proposed, where it assigns the power to all subcarriers per each NOMA user iteratively. Secondly, two power assignment schemes are proposed, namely; high weight and low weight based power assignment (PA) schemes. Where they depend on constant weights to assign the power. In particular, the high weight based PA scheme depends on a weighting factor that uses the user index within the decoding order to fractions a high portion of the available total transmission power and allocate it to the respective user. On the other hand, the low weight based PA works based on a weighting factor that fractions low amount of the total transmission power. Both of these techniques take into account each user rank index within the decoding order with respect to its counterpart, and they also take into consideration the total number of active users.

The results obtained from the simulations reveal that the proposed PA techniques maintains very good SE performances that are close to the numerical solution. It is worth mentioning that the numerical solution is obtained based on genetic algorithm. Nevertheless, the simulations results also confirm that the proposed approaches surpass the fractional transmission power approach simulated from literature works.

本文研究了通过根据每个NOMA用户的信道状态优化分配功率来提高非正交多址（NOMA）频谱效率（SE）。这将根据信道状态变化以自适应方式改善整体NOMA系统性能。提出了SE最大化的优化问题，以优化NOMA系统的性能。为了解决提出的问题，提出了三种功率分配方案。首先，提出了一种基于恒定权重的注水方法，该方法将每个NOMA用户的功率迭代分配给所有子载波。其次，提出了两种功率分配方案：高权重和低权重的功率分配（PA）方案。它们依靠恒定的权重来分配功率。特别是，基于高权重的PA方案取决于加权因子，该加权因子使用解码顺序内的用户索引来分割可用总传输功率的很大一部分，并将其分配给各个用户。另一方面，基于低权重的功率放大器基于将总传输功率的少量分成几部分的加权因子来工作。这两种技术都考虑了相对于其对应方的解码顺序内的每个用户等级索引，并且它们还考虑了活动用户的总数。基于低权重的功率放大器基于权重因子进行工作，该权重因子将总传输功率的极小部分进行分配。这两种技术都考虑了相对于其对应方的解码顺序内的每个用户等级索引，并且它们还考虑了活动用户的总数。基于低权重的功率放大器基于权重因子进行工作，该权重因子将总传输功率的极小部分进行分配。这两种技术都考虑了相对于其对应方的解码顺序内的每个用户等级索引，并且它们还考虑了活动用户的总数。

从模拟获得的结果表明，所提出的PA技术保持了非常好的SE性能，接近于数值解。值得一提的是，数值解是基于遗传算法获得的。尽管如此，仿真结果也证实了所提出的方法优于文献工作中模拟的分数传输功率方法。