We study the joint user association and resource allocation problem in both uplink (UL) and downlink (DL) for full-duplex heterogeneous cellular networks (HCNs), wherein base stations (BSs) are densely deployed with reusable subchannels and highly variable transmit powers. To reap the benefits of BS densification, decoupled multiple-association (DMA) is considered, which enables each user equipment (UE) to associate with multiple BSs for UL and DL in a decoupled manner. Furthermore, in order to provide the best service, appropriate holistic subchannel and power allocation are jointly studied and an optimization problem is formulated. However, it is challenging to solve the joint problem due to its combinatorial nature. To this end, we formulate a novel distributed four-sided matching game in which the UEs, BSs, subchannels, and power levels are ranked based on designed preference metrics for optimal matching. To obtain the solution, a low-complexity algorithm is developed. The convergence of the algorithm to a stable matching is proved and the worst-case complexity is analyzed. Numerical results are presented to demonstrate the performance of the proposed scheme in terms of the UEs’ sum-rate in UL and DL, respectively. The superiority of DMA is also investigated by comparisons.

中文翻译：

---

全双工异构蜂窝网络中的联合解耦多关联与资源分配：四边匹配博弈


我们研究全双工异构蜂窝网络（HCN）上行链路（UL）和下行链路（DL）的联合用户关联和资源分配问题，其中基站（BS）密集部署，具有可重用的子信道和高度可变的发射功率。为了获得基站致密化的好处，需要考虑解耦多关联（DMA），这使得每个用户设备（UE）能够以解耦方式与用于UL和DL的多个BS关联。此外，为了提供最佳服务，共同研究适当的整体子信道和功率分配并制定优化问题。然而，由于其组合性质，解决联合问题具有挑战性。为此，我们制定了一种新颖的分布式四边匹配游戏，其中根据设计的偏好指标对 UE、BS、子信道和功率级别进行排序，以实现最佳匹配。为了获得解决方案，开发了一种低复杂度算法。证明了该算法对稳定匹配的收敛性，并分析了最坏情况的复杂度。给出的数值结果分别证明了所提出方案在 UL 和 DL 中 UE 总速率方面的性能。通过比较也考察了DMA的优越性。