面向5G的非正交多址接入技术研究

Non-orthogonal multiple access (NOMA) allows controllable interferences by non-orthogonal resource allocation, which can efficiently improve spectral efficiency and accommodate massive connectivity. Therefore, NOMA has been considered as a promising revolutionary technology to address the challenge of a 1000-fold data traffic increase for 5G by 2020. However, there are still some challenging problems in existing NOMA schemes, e.g., incomplete theoretical analysis tool and inmature solutions. To this end, in this project we rely on graph theory, complex domain multidimensional constellation theory, and compressive sensing theory to address key challenges of NOMA in terms of basic theory and key technology. For basic theory, we consider the special structure of NOMA to obtain capacity bounds, and then the bound of overloading factor will be analyzed. For key technology, first, we will optimize the design of spreading sequences, and the receiver will also be optimized to realize the tradeoff between performance and complexity. Second, based on the compressive sensing theory, we will realize accurate user activity detection by excavating and utilizing the sparsity of user activity, and thus uplink grant-free NOMA with low overhead and low latency can be realized. Finally, by the combination of NOMA and massive MIMO, the spectral efficiency and system reliability of NOMA can be further enhanced by exploiting the space resources. Through the basic theory and key technology investigated for NOMA, this project is expected to provide strong technical support for China to lead the future 5G standard.

面对未来5G网络容量“十年千倍”的技术挑战，非正交多址接入（NOMA）通过引入可控的干扰来大幅提高频谱效率和连接数密度，成为5G备选关键技术之一，但仍面临理论体系不完善、关键技术不成熟等诸多挑战。为此，本项目以图论、复数域多维星座图理论和压缩感知理论为理论工具，深入研究面向5G的NOMA基础理论和关键技术。在理论研究方面，我们将结合NOMA特有的系统结构分析其上下行可达信道容量界，并从理论上分析系统能达到的最大过载因子。在关键技术方面，首先，实现扩频序列的优化设计和接收机算法性能与复杂度的折中；其次，挖掘并利用海量连接中自然存在的活跃用户稀疏性，提出基于压缩感知理论的随机多用户检测方法，从而实现低开销低时延的上行免调度NOMA传输机制；最后，将NOMA与Massive MIMO有机结合，进一步利用空间资源来提高频谱效率和可靠性。本项目可望为我国引领未来5G标准提供相关专利技术储备。

面对未来无线通信系统数据流量和连接数爆炸式增长的需求，非正交多址接入（Non-Orthogonal Multiple Access，NOMA）技术通过引入可控的干扰来大幅提高频谱效率和连接数密度，从而成为未来无线通信系统的关键技术之一。然而，NOMA技术面临理论体系不完善、关键技术不成熟等诸多挑战。为此，本项目基于压缩感知理论和优化理论，深入研究面向未来无线通信的NOMA基础理论和关键技术。在理论研究方面，考虑NOMA特有的系统结构，分析莱斯信道下的可达容量界，为NOMA关键技术的设计提供了理论依据。在关键技术方面，针对上行传输，挖掘并利用海量连接中自然存在的活跃用户稀疏性，基于压缩感知理论设计低开销低时延的上行免调度NOMA传输机制，并分别针对基于帧结构的传输架构和基于突发传输的架构提出相应的活跃用户与数据联合检测算法，解决上行免调度NOMA传输中活跃用户未知的难题；针对下行传输，充分利用NOMA的技术优势，提出毫米波大规模MIMO-NOMA传输机制，突破传统毫米波大规模MIMO系统用户数受限于空域自由度的基本限制，进一步，在提出的毫米波大规模MIMO-NOMA传输机制中，对预编码与功率分配进行联合优化设计，有效提高系统的谱效和能效。本项目研究成果为相关领域的研究人员提供了重要参考，将为NOMA在未来无线通信系统中的应用提供重要的理论基础和技术储备。