Distributed hybrid beamforming for cooperative millimeter-wave cell-free massive MIMO

• Main Author: Lemos, Rui Jorge Barreto Sá
• Publication Date: 2022
• Format: Master thesis
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10773/39377
• topic: Beyond 5G; Massive MIMO; Hybrid analog-digital systems; Beamforming; Cell-Free; Millimeter Waves

With the constant growing demands from users in terms of data rates, coverage and support to Internet of Things (IoT) devices, beyond 5G networks will enable a more human-friendly society with improved capacities to cope with users requirements. Cell-Free networks appear as promising concept for future wireless communications with the main goal of providing a uniform service within the coverage area. This is possible through the several distributed cooperating antennas in a geographic area, where cell-edges are eliminated thus reliving the network from inter-cell interference. Further, the integration of massive MIMO technology in the context of Cell-Free networks will bring a huge performance improvement when compared to co-located massive MIMO in traditional cell based systems. In addition, the integration of massive MIMO with the millimeter wave band requires a new architecture design to overcome the constraints of a high frequency transmission where higher path losses and a lower energy efficiency are observed. Hence, hybrid analog-digital beamforming is suggested taking advantage of both gains from analog and digital precoding reducing implementation costs, overall system complexity and improving the energy efficiency. This dissertation comprises the design and implementation of a downlink mmWave Cell-Free mMIMO with hybrid beamforming considering a distributed architecture. This architecture is characterized by splitting the data processing between the CU and the APs. Then, three scenarios were evaluated regarding the variation of the number of deployed antennas and the number of Access Point (AP)s. Also, two beamforming techniques were proposed and evaluated considering different levels of Channel State Information (CSI). The first contemplates random angles between APs and User Terminal (UT)s. The second considers the knowledge of average Angle of Departure/Arrival between APs and User Terminal (UT)s. The distributed architecture performance was compared with the centralized counterpart and the results show that the system performance is boosted when the APs are alleviated from signal processing tasks, therefore proving that the centralized architecture provides better results. Further conclusions from the results show that when the number of APs are increased, a higher diversity gain is achieved thus eliminating interference.