5G密集网络SON面临高智能、无冲突且可靠、大规模、主动性等新需求与挑战。本课题主要研究5G密集网络下多域环境感知的自组织机理与自优化方法，建立5G密集网络基于大数据分析的网络自组织机理，完善SON自优化理论。具体研究内容：（1）提出基于大数据分析的多域环境感知5G自组织网络框架，探索基于大数据分析的自组织机理；（2）提出5G密集网络下大数据分析与数据挖掘方法，构建网络自优化知识库；（3）提出基于贝叶斯网络的上下文认知模型，支持5G密集网络下主动在线学习能力；（4）提出5G密集网络基于多域环境感知的系列自优化方法，包括容量和覆盖优化、干扰自优化、移动鲁棒性自优化、移动负载均衡自优化等；（5）搭建5G密集网络自组织网络系统级仿真与验证平台，通过仿真评估与验证平台评估所提出方案的性能；本课题研究将为5G密集网络SON机理与自优化方法提供理论和技术支撑。课题具有重大的理论研究意义和实际应用价值

In 5G dense network, SON is expected to be a more intelligence SON, a conflict-free and reliable SON, massive SON and proactive SON. This project researches on the self-organization mechanism and self-optimization method in multi-domain environment aware for 5G. In order to improve self-organizing mechanisms and self-optimization theories with big data for enabling 5G. We put forward (1) the multi-domain environment-aware SON framework based on big data analysis, and explore self-organization mechanism with big data for enabling 5G; (2) big data analysis and data mining methods, and construct a self-optimizing knowledge database ; (3) the context-aware model based on Bayesian network, support active and online learning ability for 5G; (4) many self-optimizing methods with big data, such as coverage and capacity optimization, interference management, mobile robustness optimization, mobile load balancing optimization; (5) implement system level simulation platform to evaluate the performance of the proposed SON schemes with big data. This research will provide theoretical and technical support to empower self-organization mechanism and self-optimization method for enabling 5G, so our schemes have a great theoretical and practical application value.