Performance Evaluation of Low Complexity Digital Beamforming Algorithms by Link-Level Simulations and Outdoor Experimental Trials for 5G Low-SHF-Band Massive MIMO

Satoshi SUYAMA
Kohei IZUI
Kenichiro YAMAZAKI
Yukihiko OKUMURA

IEICE TRANSACTIONS on Communications   Vol.E102-B    No.8    pp.1382-1389
Publication Date: 2019/08/01
Publicized: 2019/02/20
Online ISSN: 1745-1345
DOI: 10.1587/transcom.2018TTP0022
Type of Manuscript: Special Section PAPER (Special Section on Technology Trials and Proof-of-Concept Activities for 5G and Beyond)
5G,  Massive MIMO,  digital beamforming,  outdoor experimental trials,  low SHF bands,  

Full Text: PDF(3.4MB)>>
Buy this Article

The beamforming (BF) provided by Massive MIMO is a promising technique for the fifth-generation (5G) mobile communication system. In low SHF bands such as 3-6GHz, fully digital Massive MIMO can be a feasible option. Previous works proposed eigenvector zero-forcing (E-ZF) as a digital precoding algorithm to lower the complexity of block diagonalization (BD). On the other hand, another previous work aiming to reduce complexity of BD due to the number of antenna elements proposed digital fixed BF and channel-state-information based precoding (Digital FBCP) with BD whose parameter is the number of beams. Moreover, in order to lower the complexity of the Digital FBCP with BD while retaining the transmission performance, this paper proposes Digital FBCP with E-ZF as a lower complexity digital BF algorithm. The pros and cons of these digital BF algorithms in terms of transmission performance and computational complexity are clarified to select the most appropriate algorithm for the fully digital Massive MIMO. Furthermore, E-ZF can be implemented to 4.5GHz-band fully digital Massive MIMO equipment only when the number of antenna elements is less than or equal to 64, and thus 5G experimental trial employing E-ZF was carried out in Tokyo, Japan where early 5G commercial services will launch. To the best of our knowledge, this was the first outdoor experiment on 4.5GHz-ban Massive MIMO in a dense urban area. An outdoor experiment in a rural area was also carried out. This paper shows both a coverage performance under the single user condition and system throughput performance under a densely deployed four-user condition in the outdoor experimental trials employing the E-ZF algorithm. We reveal that, in the MU-MIMO experiment, the measured system throughput is almost 80% of the maximum system throughput even if users are closely located in the dense urban area thanks to the E-ZF algorithm.