基于无线通信大数据的传输理论与技术

Recent years we have witnessed the explosion of the wireless communication user services as well as the system scale and complexity. New services and data characteristics have imposed significant challenges to the design and optimization of the future wireless communication systems. The objective of this project is to explore the wireless communication big data sets originated from different sources, and combine general and specific data mining and analysis methodologies to study wireless transmission theories and technologies. Specifically, we investigate the dynamic correlations among massive users, base station deployments, services and spectra. We then construct the big data feature bases for the development of transmission theories and technologies, and further reveal the mechanisms for the massive users diverse services across different dimensions. We further develop the information theories on the clustered users and dense base stations, and propose the source coding for the clustered users, the non-orthogonal transmissions, the transmission methodologies that match the big-data interference characteristics, and the full-spectrum fused transmission mechanism. The research results may demonstrate the system performance improvements brought by the big data analysis. We also construct the transmission technology verification platform and database sharing platform suitable for the wireless big data communication environment. It is believed that this project may lay a foundation for the big data theories and technologies for the future green wireless communications, and build up a large-scale integrated research and development environment for the big data-based wireless communications.

无线通信无论从用户业务还是系统规模与复杂度方面，近几年均呈现显著增长态势，相应的新型业务与数据特征对未来无线通信系统设计与优化提出了严峻挑战。本项目基于多源无线通信大数据集合，采用通用与特色相结合的大数据分析方法，挖掘面向无线通信的用户站点和业务频谱的动态关联特征，构建服务于传输理论与技术研究的大数据特征集，进一步揭示巨量用户多样化业务的跨维度分布规律。并且探索群聚用户和密集站点覆盖下的信息理论，提出群聚编码和非正交传输技术，以及大数据干扰下特征匹配传输方法，阐明全频谱融合传输机制，展示无线通信大数据增益，并构建无线通信大数据环境下的传输方法验证平台与数据库共享平台，为未来无线绿色通信奠定大数据理论与技术基础，并营造规模化和集成化无线通信大数据开放研究与应用环境。

本项目针对混杂密集干扰下群聚用户的信道容量及编码理论、基于大数据业务特征的密集覆盖传输理论方法、群聚用户多样化业务驱动的全频谱融合传输机制这三个关键科学问题，进行了深入研究，研究成果包括但不限于：密集干扰下的无线光通信单位面积传输速率、基于干扰地图的密集干扰无线光通信编码技术、基于知识驱动机器学习的无线信道估计、无线用户行为的预测理论及应用、机器学习辅助的大规模MIMO无线信道预测、免调度大规模异步无线接入下的高效接收方法、基于数据驱动的视距链路传输以及单次散射链路传输关键技术。项目团队搭建了室内宽光谱无线光通信密集用户接入实验平台，宽光谱室内信道与通信测试平台，并构建了无线大数据共享平台。这些成果为未来无线通信大数据系统的设计与优化提供了传输理论与关键技术方面的重要支撑。