基于多维非正交扰动的大规模MIMO-OFDM非线性失真抑制技术

Nonlinear distortion mitigation for massive MIMO-OFDM downlink has been regarded as one of the main bottlenecks constraining the realization of massive MIMO-OFDM systems. The existing nonlinear distortion mitigation schemes do not exploit the freedom offered by massive MIMO-OFDM systems, and ignore the problem that nonlinear distortion cannot be measured effectively only by PAPR. Therefore, they cannot achieve the optimal overall performance. To overcome the drawbacks, the project first builds a general signal perturbation model with the concept of multi-dimensional perturbation space. Then, based on the general signal perturbation model, the project illuminates the mechanism of nonlinear distortion mitigation with multi-dimensional perturbation. Thirdly, the project proposes optimized multi-dimensional non-orthogonal perturbation based nonlinear distortion mitigation schemes with multiple metrics. Finally, optimization methods are developed with low complexity to accomplish effective nonlinear distortion mitigation. Through the basic research and key technologies investigated for massive MIMO-OFDM systems, we can expect to both improve the system performance and lower the energy cost.

大规模MIMO-OFDM系统下行链路非线性失真抑制问题是大规模MIMO-OFDM系统走向实用化的主要瓶颈之一。针对现有大规模MIMO-OFDM非线性失真抑制技术对系统的自由度利用不充分及使用单一的峰均比准则无法有效衡量非线性失真对系统性能影响的问题，本课题首先提出多维扰动空间的概念，通过建立通用信号扰动模型，分析不同扰动空间所具有的扰动自由度，来揭示基于多维空间扰动的大规模MIMO-OFDM非线性失真抑制机理；通过设计多准则约束下的多维非正交扰动非线性失真抑制优化方案和低复杂度的优化算法，来实现有效的非线性失真抑制。本研究不仅可以有效降低大规模MIMO-OFDM系统由于非线性失真所带来的性能损失问题，而且可以有效提升发端设备功率效率，为构建下一代大容量、高可靠的无线通信网的实用化提供必要的理论及技术支撑。

大规模MIMO-OFDM系统下行链路非线性失真是制约大规模MIMO-OFDM系统走向实用化的主要瓶颈之一。本项目从多维扰动空间的角度出发，深度挖掘了大规模MIMO-OFDM系统数据载波扰动空间、时域扰动空间、预编码矩阵扰动空间和空闲载波扰动空间的扰动自由度及其对系统性能指标的影响。在此基础上，针对数据载波扰动空间设计了一种具有更多自由度的结构化扩展星座，并提出了初始概率自适应的增强型交叉熵载波注入非线性失真抑制方案。针对时域扰动空间，提出了一种广义分段线性压扩非线性失真抑制方案。最后，从系统性能整体最优出发，设计了多准则约束下的多维扰动非线性失真抑制优化方案及低复杂度的优化求解算法，实现了非线性失真的有效抑制。本项目有的研究成果对应有效降低大规模MIMO-OFDM系统非线性失真，提升系统性能具有理论指导意义。