基于混合全/半双工的5G超密集网络中无线资源管理技术研究

In order to adapt to the exponential increase in wireless devices and mobile data, service providers, manufacturers and standard organizations all over the world have paid a lot of attentions to the spectral efficiency and capacity enhancement technologies in next generation (5G) cellular networks. Full-duplex wireless and hyper-dense networks are two promising technologies to improve the spectrum efficiency and capacity of 5G wireless networks. Although there are a large number of exiting works focusing on full-duplex wireless and hyper-dense networks respectively, most of the papers on hyper-dense networks assume that base stations can only operate in half-duplex mode. Since the researches on hyper-dense networks with full-duplex wireless is still at its infancy, significant challenges remain to be addressed. In order to take the advantages of full-duplex technologies while considering the compatibility to the existing half-duplex networks , we focus on the resource management techniques in hyper-dense networks with hybrid full/half duplex mode. The main tasks are listed as follows: 1) To dynamically switch between full and half duplex mode at each base station, we will investigate the duplex mode switching scheme for 5G hyper-dense networks based on stochastic geometry; 2)To mitigate the inter-user interference, the joint user association and power allocation algorithm will also be addressed using geometric programming; 3) To utilize the radio resource efficiently and improve the network throughput, we will also study the joint uplink and downlink resource allocation problem while guaranteeing the statistical quality of service.

为适应设备数量和数据流量的爆炸式增长，各国运营商、设备商及标准组织已在积极研究下一代（5G）网络中的频谱效率和容量增强技术。全双工无线通信和超密集网络均是提升5G网络频谱效率和容量的潜在关键技术。虽然已有大量文献分别针对全双工无线通信和超密集网络展开了深入研究，但在超密集网络相关研究中，多数假设基站工作在半双工模式下，关于全双工通信在超密集网络中的应用研究仍处于初期阶段，还存在大量挑战和机遇。为兼容现有半双工网络并充分利用全双工技术的优势，本课题考虑基于混合全/半双工的5G超密集网络，研究其无线资源管理技术，主要研究内容包括：第一，基于随机几何理论研究基站全/半双工模式选择机制，以实现基站双工模式的动态切换；第二，利用几何规划理论设计联合用户关联与功率控制算法，以有效应对用户间干扰；第三，在保证统计QoS的前提下，研究上下行资源联合分配算法，以高效利用上下行无线资源，提高网络的吞吐量。

按照原定计划，本课题重点针对接入设备双工模式的动态切换、功率控制算法以及统计QoS保障机制等展开研究，完成了原定研究计划。在此基础上，进一步考虑了非正交多址接入技术与全双工无线通信技术的结合，考虑5G新兴技术（车联网等）的基础上对项目研究内容进行了一定的拓展。具体研究了如下内容：. ①混合全/半双工非正交多址接入系统中的最优功率分配策略. ②基于统计QoS理论提出了非正交多址接入系统中的最优功率分配策略. ③基于直传链路与中继链路协作的非正交多址接入系统中功率分配问题. ④低时延高可靠V2X网络协作非正交广播/多播方法. 基于上述研究内容，以第一作者或通信作者发表重要学术期刊或会议论文13篇，其中SCI论文5篇，EI论文8篇；其中以第一作者发表于IEEE Trans. Wireless Communication的论文在2018年入选了ESI高被引论文，获得了国内外众多知名学者的证明引用与评述，为混合全/半双工非正交多址接入系统中双工模式切换和功率分配问题奠定了理论基础。