Spatial Compatibility Metrics Applied in 5G C-RAN Networks

Abstract

Network densification, millimiter wave (mmWave) spectrum frequencies, and massive multiple- input multiple-output (MIMO) are three key technologies pointed out by the research community and industry to meet the system capacity requirements for the fifth generation (5G) of wireless communications and beyond fifth generation (B5G). However, some challenges emerge during the application of these technologies. In this thesis, we investigate strategies to handle some problems inherent to these technologies by using spatial compatibility metrics that map properties of MIMO channels to evaluate how efficiently such channels can be separated in space. This thesis is divided into two parts. The first part aims to handle interference in a conventional MIMO small-cell network. The network densification leads to fast fluctuation and unbalanced traffic demand between uplink and downlink due to the small number of user equipments (UEs) in each cell. Dynamic time division duplex (DTDD), where each base station (BS) chooses its transmission direction, has been considered as a promising solution for this issue. In such scenario, cross-interference is created between BSs and UEs. To manage this interference, a spatial compatibility metric is proposed based on two parameters that control the trade-off between intended channel attenuation, cross-channel correlation, and co-channel correlation. The proposed spatial metric was evaluated in a multi-cell scheduling problem that was solved by using different optimization techniques.. In the second part, the focus is on hybrid beamforming (HBF) multi-cell scenario operating in mmWave frequencies. In such technology, the antenna array of BSs are connected to a smaller number of radio frequency (RF) chains to reduce costs and power consumption. In such architectures, the beamforming is separated into analog beamforming and a digital precoder. In this thesis, a greedy algorithm based on a spatial compatibility metric is proposed for analog-beam assignment. The last proposal of this thesis is a UE-BS association and HBF design framework. The UE-BS association is based on the spatial similarity of the channel to avoid inter-cell interference. The optimization problems for sum-rate maximization and power minimization have also been considered. The proposed framework is shown to achieve a good trade-off between energy efficiency (EE) and feasibility of solutions.