大规模超密集异构云无线接入网高效传输关键技术研究

The centralization of baseband processing in large-scale and ultra-dense heterogeneous cloud radio access network (H-CRAN) can achieve the overall optimization of signal processing and radio resource, which leads to a significant improvement of transmission performance. However, massive access and large dimensional resource allocation increase the burden of network management. In order to satisfy the new requirement such as massive access, high spectrum efficiency and high energy efficiency in future wireless communication systems, this project studies the key techniques of low-complexity and high-efficient transmission in large-scale and ultra-dense H-CRAN. The main contents of the project include the following: (1) By excavating the sparsity of user activity and user signals, efficient algorithms for active user detection and channel estimation are developed based on deep learning. (2) Through considering the power consumption of the overall system, low complexity resource allocation methods based on first-order algorithm are designed. By adopting the frontier theory such as deep learning, compressed sensing and non-convex optimization, the research results of this project will significantly increase spectrum efficiency and energy efficiency of large-scale and ultra-dense H-CRAN.

大规模超密集异构云无线接入网（Heterogeneous Cloud Radio Access Network，H-CRAN）基带处理的集中化能够实现信号处理和无线资源的全局优化配置，实现网络传输能力的阶跃式提升。然而，海量终端接入和大维资源优化加重了网络管理负担。为了满足未来无线通信系统的海量接入、高谱效、高能效等新需求，本项目研究大规模超密集H-CRAN的低复杂度高效传输关键技术。主要内容包括：（1）挖掘用户活跃的稀疏性和用户信号的空间稀疏性，提出基于深度学习的高效活跃用户检测与信道估计方法；（2）综合考虑系统功耗，设计基于一阶算法的低复杂度资源分配方法。通过采用深度学习、压缩感知和非凸优化等前沿理论，本项目的研究成果将显著提高大规模超密集H-CRAN中的频谱效率和能量效率。

大规模超密集异构云无线接入网（Heterogeneous Cloud Radio Access Network，H-CRAN）基带处理的集中化能够实现信号处理和无线资源的全局优化配置，实现网络传输能力的阶跃式提升。为了满足未来无线通信系统的海量接入、高谱效、高能效等新需求，本项目研究大规模超密集H-CRAN的低复杂度高效传输关键技术。主要内容包括：1）研究大规模超密集H-CRAN中的活跃用户检测与信道估计算法，提出了一种基于稀疏界学习的自适应权重加权算法，仿真显示：相比传统算法，所提算法在稀疏度、压缩比和检测概率方面可以获得更好的性能。2）研究了大规模超密集H-CRAN中的低复杂度资源分配算法，提出了一种一阶算法及其加速算法，仿真显示：所提的一阶算法可以以非常低的计算复杂度获得与传统的内点法差不多的性能。本项目研究了大规模超密集H-CRAN中高效传输技术，为面向海量接入的无线接入网中高效资源分配提供了理论依据和技术参考，对H-CAN的实用化具有重要意义。