
China Communications • February 2019
95
With the fast development of wireless devices,
the data traffic has exponentially increased
for wireless networks. In order to meet future
traffic demands, academia and industry are
committed to developing the next genera-
tion of mobile cellular communication (5G).
Therefore, millimeter wave (mmWave) fre-
quency bands, ranging from 30 GHz to 300
GHz, have been considered as a promising
candidate [1], [2]. Comparing with current
cellular bands, mmWave signals experience
severe propagation loss. Due to the small
wavelength of mmWave, it is feasible to pack
more antennas in the same physical dimension
to form the large-scale antenna arrays. While
the large-scale antenna arrays can provide sig-
nificant beamforming gains to overcome the
propagation loss [3]. Moreover, it also enables
the transmission of multiple data streams re-
sulting in signicant improvements to spectral
efciency.
In traditional multiple-input multiple-output
(MIMO) systems, the full-digital precoding is
completely realized in the digital domain by
operating both the magnitude and phase of the
baseband signals. However, this full-digital
scheme is impractical for mmWave MIMO
systems, due to the high cost and power con-
sumption of RF chains. To tackle the above is-
Abstract: Hybrid precoding and combining
have been considered as a promising tech-
nology, which can provide a compromise
between hardware complexity and system
performance in millimeter wave multiple-in-
put multiple-output systems. However, most
existing hybrid precoder and combiner designs
generally assume that innite resolution phase
shifters (PSs) are used to produce the analog
beamformers. In a practical scene, the design
with accurate PSs can lead to high hardware
cost and power consumption. In this paper, we
investigate the hybrid precoder and combiner
design with nite resolution PSs in millimeter
wave systems. We employ alternate optimi-
zation as the main strategy to jointly design
analog precoder and combiner. In addition, we
propose a low complexity algorithm, where
the analog beamformers are implemented only
by nite resolution PSs to maximize spectral
efficiency. Then, the digital precoder and
combiner are designed based on the obtained
analog beamformers to improve the spectral
efficiency. Finally, simulation results and
mathematical analysis show that the proposed
algorithm with low-resolution PSs can achieve
near-optimal performance and have low com-
plexity.
Keywords: millimeter wave; multiple-input
multiple-output; finite resolution phase shift-
ers; hybrid precoding; alternate optimization
Received: Sep. 4, 2018
Revised: Dec. 6, 2018
Editor: Kai Kang
Ran Zhang
1
, Weixia Zou
1,2,
*
, Ye Wang
1
, Mingyang Cui
1
1
Key Lab of Universal Wireless Communications, MOE, Beijing University of Posts and Telecommunications, Beijing 100876, China
2
State Key Lab. of Millimeter Waves, Southeast University, Nanjing 210096, China
* The corresponding author, email: zwx0218@bupt.edu.cn
MILLIMETER AND THZ WAVE FOR 5G AND BEYOND