Simulation and Testing of a MU-MIMO Beamforming System

Sammanfattning: Multi-User Multiple-Input Multiple-Output (MU-MIMO) technology has become increasingly important in the field of wireless communication due to its ability to highly increase the capacity and efficiency of wireless networks [1]. Beamforming, as a technique used in MU-MIMO systems, improves network performance by improving signal quality and reducing interference. With the emergence of 5G and beyond, the complexity of Advanced Antenna Systems (AASs) that perform beamforming has increased considerably. Consequently, testing the AAS before installation becomes vital to ensure the reliability and performance of the system. Meanwhile, the Butler matrix has gained significant attention as a passive device for efficient and cost-effective testing of the AAS and beamforming setup. Generally, the Butler matrix can be used in the Base Station (BS) of mobile networks to create beams towards the User Equipment (UEs). In this thesis work, a method for beamforming tests based on channel reciprocity in MU-MIMO is studied. A beamforming setup in the laboratory using a Butler matrix to form beams is used before the BS is deployed in real-world scenarios. Based on the Sounding Reference Signal (SRS), which is received from the UEs, the BS estimates the channel for each UE separately and applies the appropriate weight matrix to determine beams towards the UEs. The purpose of the study is to evaluate and validate the system for beamforming tests. Assessment is carried out in two parts. Taking everything into account, we first simulate a combination of directional signals using an Over The Air (OTA) test method. This simulation involves four UE positions. This approach enables us to verify the accuracy of the beam patterns generated by the system. Furthermore, it identifies side lobes that might be present in the beam patterns. Through these simulations, we can mitigate and reduce these side lobes, enhancing the overall quality of the testing process. Secondly, measurements were performed in the Downlink (DL) and Uplink (UL) modes. In the DL measurements, the Physical Downlink Shared Channel (PDSCH) power and throughput of the UEs were measured for the setup in the laboratory. Subsequently, on the basis of our observation of low PDSCH power or low throughput values, we explored the root causes using UL measurements. The UL measurements involved recording the SRS traces. Our approach for obtaining SRS data, testing methodologies, precise data selection from log traces, and subsequently beam mapping algorithm are described in detail. The study includes a comprehensive description of the methodology along with the corresponding results, ensuring a complete understanding of the process.