密集异构通信网基于信道对角块化的干扰对齐设计方法研究

As an effective interference management technique, interference alignment (IA) can solve the interference problem in ultra-dense heterogeneous mobile communication networks. However, the traditional IA design methods cannot achieve the highest system performance by combining space, time and frequency resources. This project is to propose the IA design method based on diagonal-block channel character, which aims to maximize the system sum-rate performance in ultra-dense heterogeneous networks, by exploiting multiple radio resources fully. Firstly, to judge the feasibility of IA design based on diagonal-block channel character, this project is to derive the necessary conditions of IA feasibility by using results in algebraic geometry, and study the feasibility conditions from the angle of diagonal-block channel matrix dimension. Secondly, by using the results in matrix theory, this project is to study the specific IA solutions with low computational complexity based on diagonal-block matrix character, which can achieve high system performance. Finally, the diagonal-block channel error model is established for the imperfect channel state information case. Furthermore, by using the results in probability theory and information theory, this project is to quantify the IA system performance loss theory, and study the impact of diagonal-block channel character on IA system performance loss. The IA design method based on diagonal-block channel character is expected to improve the system performance in ultra-dense heterogeneous networks, which has important theoretical significance and practical value.

干扰对齐是解决移动通信超密集异构网络中干扰问题的一项有效干扰管理技术，但目前结合空、时、频多域资源传统的干扰对齐设计方法无法实现最高系统性能。针对这一问题，本项目拟提出基于信道对角块化的干扰对齐设计方法，旨在通过充分利用多域无线资源，最大化提高超密集异构网络的系统和速率性能：首先，针对信道对角块化的干扰对齐设计可行性判断，利用代数几何推导出其可行的必要条件，并从对角块信道矩阵维度的角度研究其可行性条件；其次，利用矩阵理论，研究实现高系统性能、低运算复杂度的信道对角块化干扰对齐具体求解方法；最后，针对非理想信道参数，建立对角块信道误差模型，利用概率论和信息论量化干扰对齐系统性能损失，并研究信道对角块化对其系统性能损失的影响。基于信道对角块化的干扰对齐设计方法有望提高超密集异构网络的系统性能，具有重要的理论和应用价值。

随着移动通信系统的发展，超密集异构组网使得蜂窝通信网络变的更加异构、复杂，导致同频干扰问题愈加严重。干扰对齐作为一种有效的干扰管理技术，能够解决未来通信网络的干扰问题。但面向实际通信网络应用干扰对齐技术还面临一些关键理论与技术问题，如充分结合空、时、频多域资源如何实现最高和速率系统性能。针对这一问题，本项目以超密集异构网络为背景，将信道对角块化为切入点，研究了实现最高和速率系统性能的干扰对齐设计方法，并取得了一批较好的成果：1）首先，建立了多种干扰网络模型下干扰对齐通信系统模型，利用代数几何提出了干扰对齐可行的必要条件与充分条件，并构建了自由度增益基础理论；2）然后，以自由度性能指标为指导，利用矩阵运算理论与方法论提出了低复杂度的干扰对齐具体求解方法，包括具有快速收敛性的迭代算法和基于对齐链的闭式解求解方法，其中基于对齐链的闭式解求解方法具有一定创新性，有效推动了干扰对齐闭式解问题的研究进展；3）最后，针对非理想信道状态信息情况，利用概率论量化了干扰对齐系统和速率性能损失，并研究了非理想信道参数对其系统性能损失的影响。这些研究成果为后续深化和拓展相关领域的研究工作（如缓存辅助无线干扰网络、智能超表面辅助通信网络中的干扰管理）奠定了良好基础，对于充分利用多域资源提高系统性能具有重要的理论意义与应用价值。