超密集分层异构网络能效增强研究

To tackle the 1000× mobile data challenge, one clear enabler is Ultra-dense Heterogeneous Networks (HetNets) deployed with a mix of traditional macrocells and Low Power Nodes (LPNs) such as pico, femto, and/or relay nodes. However, ultra-dense deployments of LPNs in HetNets will cause issues of network energy consumption increase and more complex inter-tier and intra-tier interference，which will severely deteriorate the network energy efficiency (EE) obviously. The current studies for HetNets EE enhancement mainly focus on the network scenario with LPNs sparse deployment, which cannot tackle the new challenges caused by LPNs ultra-dense deployment. To solve these troubles, we will adopt Poison Point Process (PPP) to model the network scenario, then we will conduct research in three levels: the Base Station (BS) sleep/awake metric, the inter-tier and intra-tier interference suppression, and the trade-off between EE and spectrum efficiency (SE). We will firstly propose BS adaptive sleep/awake period optimization algorithm based on the combination analysis of large scale time statistics of cell load fluctuation and small scale time of cell load dynamic change. Next, based on time domain interference coordination, we will discuss the Almost Blank Subframe (ABS) pattern configuration strategies for interference repression among adjacent BSs. Finally, the trade-off between EE and SE will be investigated based on time-frequency resource allocation on different kinds of subframes.

为应对5G蜂窝网络1000倍网络数据容量增长挑战，由宏蜂窝与超密集部署小基站混合同频组网的超密集分层异构网络，可极大幅度的提升蜂窝网络容量，成为5G的关键候选技术。然而超密集部署小基站会带来网络能耗增加和更加复杂的网络层间和层内干扰，造成网络能效的显著下降。当前针对分层异构网络节能的研究主要关注于小基站的稀疏部署场景，不能有效应对小基站超密集部署带来的新挑战。为此，本项目将基于泊松点过程对网络架构建模，从基站自适应睡眠/唤醒机制，网络层间和层内干扰的联合抑制方法，能量效率和频谱效率联合优化的时频资源分配设计三个层次上开展研究。并对联合长时间业务负载变化统计规律和短时间业务动态变化的基站睡眠及唤醒周期的优化，层间及层内干扰基站间基于时域干扰协调的几乎空子帧模式配置，基于不同子帧类型的能量效率和频谱效率折中准则设计等关键科学问题进行重点攻关，为未来5G绿色蜂窝通信提供有力的解决方案。

针对高能效基站自适应睡眠/唤醒机制，根据网络负载状况，提出了一种联合长时间业务负载变化统计规律和短时间业务动态变化的小基站动态睡眠/唤醒机制，在实现小基站节能的同时，兼顾小基站有偶然性活动用户出现时的服务等待时间；网络层间和层内干扰的联合抑制方面，基于随机几何理论对网络场景进行建模，联合研究小基站部署密度优化以及增强型小区间干扰协调技术，实现分层异构网络的能效最大化；高能效时频资源分配算法研究方面，针对Macro-Pico网络场景中，由于CRE技术的引入，将接入用户分为三大类，即Macro用户，Pico中心用户以及CRE用户，提出针对不同用户类别的资源调度算法，消除层间干扰的同时，优化网络能效。通过上述三个层次的研究，循序渐进的提出了一套完整的增强分层异构网络能效的算法和机制，有效改善了分层异构网络能效。