The limited fronthaul capacity is known to be one of the main problems in cloud radio access networks (CRANs), especially in the wireless fronthaul links. In this paper, we consider the uplink of a CRAN system, where massive multiple-input multiple-output (MIMO) is utilized in the fronthaul link. Considering multi-antenna user equipment (UEs) and multi-antenna remote radio heads (RRHs), we maximize the system sum-rate by jointly optimizing the precoders at the UEs and the quantization noise covariance matrices and transmit powers at the RRHs. To solve the resulting nonconvex problem, an iterative algorithm based on the majorization-minimization (MM) method is proposed. Two schemes at the central unit are considered, namely maximum ratio (MR) and zero-forcing (ZF) combining. Numerical results show that the sum-rate has an asymptotic behaviour with respect to the maximum available power at RRHs and that the MR scheme goes to its asymptote faster than the ZF scheme.

翻译：众所周知,前台容量有限是云层无线电接入网络(CRANs)中的主要问题之一,特别是在无线前台链接中。在本文中,我们考虑了CRAN系统的上链,该系统在前台链接中使用了大量的多投入多输出数据(MIMO),考虑到多亚氮用户设备(UES)和多亚氮远程无线电头(RRHs),我们通过联合优化UES和定量噪声共变矩阵的预译器和RRHs的传输能力,最大限度地实现系统总和率。为了解决由此产生的非convex问题,我们提出了基于主要化-最小化(MMM)方法的迭代算法。中央单位的两个方案得到了考虑,即最大比率(MR)和零压力(ZF)组合。数字结果显示,在RRHs现有最大功率方面,总比率具有一种无症状行为,而MRRRRHs计划也比ZF计划快。