Distributed antenna systems in 5G mobile communications

Abstract

The needs of mobile communication users increase from year to year, as there are occurring more and more applications using mobile networks. On the one side the demand for high data rates increases, while on the other side reliable low-latency connections are gaining interest, especially in public transport and in the car industry, e. g. autonomous driving. To provide these demands the capacity of future beyond 5th generation (5G) networks has to increase. This is done by using higher carrier frequencies in the range of 3-300 GHz, making use of a lot of unoccupied bandwidth. As this comes with the drawback of higher path loss and worse channel conditions the coverage of beyond 5G networks needs to improve. A possible way to do this is the deployment of distributed antenna systems (DAS). These systems make use of spatially distributed antennas, which are connected with a physical low-latency link to a central unit (CU). This CU is responsible for the coordination of the antennas and does all the signal processing. This thesis analyses DASs regarding various topics. The thesis starts with the development of a single-user DAS and the used hybrid beamforming is investigated. This includes user grouping based on channel characteristics with a threshold test and a specific grouping algorithm, used for the beamforming. The first part ends with a comparison of a DAS and a macro cell. In the second part, the system model is extended to a multi-user system and the following topics are analyzed. First, the variation of general parameters like the number of antenna array elements and radio frequency (RF) chains are analyzed. After that, a realistic per RF chain power constraint is investigated. Then channel estimation errors are introduced and compared to a codebook-based beamforming strategy and the needed number of feedback bits. Thereafter scheduling aspects are covered. The thesis ends with a comparison of a multi-user DAS to small cells. All mentioned aspects were implemented and simulated with the Vienna 5G System Level Simulator.