In this paper, the sum Spectral Efficiency (SE) maximization in a two-tier non-terrestrial heterogeneous network is investigated. This network consists of a High Altitude Aerial Base Station (HABS) and several Low Altitude Aerial Base Stations (LABSs) that share downlink spectrum with Orthogonal Frequency Division Multiple Access (OFDMA). The Aerial Base Stations (ABSs) are horizontally located based on spatial regularity maximization algorithm to maximize their coverage. Then, the weighted minimum distance criterion is selected for the users-ABS (UE-ABS) association. Due to the advantages of distributed optimization, problem formulation is performed in a distributed manner, indicating that the resource allocation is accomplished locally for each ABS. Two types of problems are formed for HABS and LABSs, respectively. The main objective of the problems is the sum SE under maximum power and inter-tier interference constraints. The defined problems are mixed continuous and combinational, where the tight upper bound solutions are derived by employing the Lagrangian Dual Decomposition (LDD) method. Finally, the effects of increasing the number of LABSs on achieving better SE performance in various cases for inter-tier interference and the effect of the weighting factor in UE-ABS association are evaluated

本文研究了两层非地面异构网络中的总频谱效率（SE）最大化。该网络由一个高空空中基站（HABS）和几个低空空中基站（LABS）组成，它们与正交频分多址（OFDMA）共享下行链路频谱。空中基站（ABS）基于空间规律性最大化算法而水平放置，以最大化其覆盖范围。然后，为用户-ABS（UE-ABS）关联选择加权的最小距离标准。由于分布式优化的优势，问题制定以分布式方式执行，这表明资源分配是在每个ABS本地完成的。HABS和LABS分别形成两种类型的问题。问题的主要目的是最大功率和层间干扰约束下的总和SE。定义的问题是混合的连续问题和组合问题，其中严格的上限解是通过使用拉格朗日对偶分解（LDD）方法得出的。最后，评估了在层间干扰的各种情况下增加LABS数量对实现更好的SE性能的影响以及UE-ABS关联中加权因子的影响