Performance Evaluation of Downlink Multi-User Massive MIMO with Configurable Active Antenna System and Inter Access Point Coordination in Low-SHF-Band

Yi JIANG  Kenichiro YAMAZAKI  Toshihiro HAYATA  Kohei IZUI  Kanada NAKAYASU  Toshifumi SATO  Tatsuki OKUYAMA  Jun MASHINO  Satoshi SUYAMA  Yukihiko OKUMURA  

Publication
IEICE TRANSACTIONS on Communications   Vol.E102-B   No.8   pp.1401-1410
Publication Date: 2019/08/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.2018TTP0013
Type of Manuscript: Special Section PAPER (Special Section on Technology Trials and Proof-of-Concept Activities for 5G and Beyond)
Category: 
Keyword: 
5G,  Massive MIMO,  low-SHF-band,  MU-MIMO,  beamforming,  C-RAN,  CB,  CoMP,  

Full Text: FreePDF(5.3MB)


Summary: 
Massive multiple input and multiple output (Massive MIMO) is a key technique to achieve high system capacity and user data rate for the fifth generation (5G) radio access network (RAN). To implement Massive MIMO in 5G, how much Massive MIMO meets our expectation with various user equipment (UEs) in different environments should be carefully addressed. We focused on using Massive MIMO in the low super-high-frequency (SHF) band, which is expected to be used for 5G commercial bands relatively soon. We previously developed a prototype low-SHF-band centralized-RAN Massive MIMO system that has a flexible active antenna system (AAS)-unit configuration and facilitates advanced radio coordination features, such as coordinated beamforming (CB) coordinated multi-point (CoMP). In this study, we conduct field trials to evaluate downlink (DL) multi-user (MU)-MIMO performance by using our prototype system in outdoor and indoor environments. The results indicate that about 96% of the maximum total DL system throughput can be achieved with 1 AAS unit outdoors and 2 AAS units indoors. We also investigate channel capacity based on the real propagation channel estimation data measured by the prototype system. Compared with without-CB mode, the channel capacity of with-CB mode increases by a maximum of 80% and 104%, respectively, when the location of UEs are randomly selected in the outdoor and indoor environments. Furthermore, the results from the field trial of with-CB mode with eight UEs indicate that the total DL system throughput and user data rate can be significantly improved.