Visible light communication for Vehicle to Everything
beyond 1 Gb/s based on an LED car headlight
and a 2 × 2 PIN array
Chaofan Wang (王超凡), Guoqiang Li (李国强), Fangchen Hu (胡昉辰),
Yiheng Zhao (赵一衡), Junlian Jia (贾俊连), Peng Zou (邹 鹏), Qiuyi Lu (卢秋仪),
Jiang Chen (陈 将), Zhongya Li (李忠亚), and Nan Chi (迟 楠)*
Academy for Engineering and Technology, Shanghai Institute for Advanced Communication and Data Science,
Key Laboratory for Information Science of Electromagnetic Waves (MoE), Fudan University, Shanghai 200433, China
*Corresponding author: nanchi@fudan.edu.cn
Received July 3, 2020; accepted July 23, 2020; posted online September 21, 2020
Visible light communication (VLC) shows great potential in Internet of Vehicle applications. A single-input
multi-output VLC system for Vehicle to Everything is proposed and demonstrated. A commercial car headlight
is used as transmitter. With a self-designed 2 × 2 positive-intrinsic-negative (PIN) array, four independent sig-
nals are received and equalized by deep-neural-network post-equalizers, respectively. Maximum-ratio combining
brings high signal-to-noise ratio and data rate gain. The transmission data rate reaches 1.25 Gb/s at 1 m and
exceeds 1 Gb/s at 4 m. To the best of our knowledge, it is the first-time demonstration of beyond 1 Gb/s employ-
ing a commercial car headlight.
Keywords: visible light communication; Internet of Vehicle; Vehicle to Everything; single-input multi-output;
deep neural networks; maximum-ratio combining.
doi: 10.3788/COL202018.110602.
Since 5G technology began to spread, Inter net of Vehicle
(IoV) has become one of the application direction s that
attracts much attention. Not only establishing connec-
tions between vehicles, but also interacting vehicles with
other terminals, Vehicle to Everything (V2X) will be the
focus of future research. Application scenarios such as
real-time traffic monitoring, autopilot, and vehicle posi-
tioning
[1–3]
demand high speed, low latency, high stability,
and high reliability. Visible light communication (VLC)
provides high-capacity, high-speed wireless data links
[4–8]
,
which is one of the powerful communication methods for
the future of IoV. As a matter of fact, most of the head-
lights of cars nowadays utilize light emitting diodes
(LEDs) as illuminators, which builds a bridge between
VLC and IoV. Combining illumination and communica-
tion, vehicle VLC meets energy-saving requirements
and has great research value.
In past research on VLC for V2X, the vast majority
involved single-input single-output (SISO) systems
[1,2,9–11]
.
However, car headlights have dedicated optical structure
designs with large divergence angles, which bring about
challenges to maintain high signal-to-noise ratio (SNR)
of the signal after long-distance transmission. Therefore,
the application of multiple receivers is an inevitable re-
quirement to increase the transmission data rate and dis-
tance. Furthermore, multiple receivers can effectively
improve the robustness of VLC for V2X systems in prac-
tical applications. Single-input multi-output (SIMO) and
multi-input multi-output (MIMO) systems are the devel-
opment trends of VLC for V2X.
Besides the dedicated optical structure, car headlights
have several characteristics that are different from LEDs
for communication. One of them is that the LEDs used in
car headlights have a relatively narrow bandwidth com-
pared to the ones used in other VLC systems. In order
to achieve high-transmission data rates under this circum-
stance, high spectral efficiency digital signal processing al-
gorithms are necessary. Over the years, the data rate of
V2X systems have gradually increased due to the progress
of algorithms
[1–3]
. Another feature is the high working
power of car headlights, which induces significant nonlin-
ear distortion of the VLC signal. For future long-distance
VLC for V2X systems, it is inevitable the signal amplitude
will be raised to maintain the needed SNR. Nonlinear dis-
tortion must be dealt with. Meanwhile, machine learning
has been a hot research area in recent years, and the ap-
plication of deep neural networks (DNNs) in communica-
tion algorithms shows a good anti-nonlinear effect
[12]
.DNN
has bright prospects in V2X systems.
In this Letter, we propose an SIMO-DNN VLC system
for V2X. The transmitting illuminator is a commercial car
headlight (Shanghai Xiao Fu Company, IP32) with an
optical structure that meets GB25991 light distribution
requirements of automotive LED headlamps. A self-
designed 2 × 2 positive-intrinsic-negative (PIN) array
receiver is used to implement the SIMO system. The 2 ×
2 PIN array receiver outputs four independent channels,
and each channel goes through a DNN post-equalizer.
Maximum ratio combining (MRC) of the four channels
results in higher SNR, which in turn results in better bit
error rate (BER) performance. The transmission distance
varies from 1 m to 5 m. By utilizing a bit-loading discrete
multi-tone (DMT) modulation scheme, the data rate
reaches 1.25 Gb/s at the transmission distance of 1 m,
COL 18(11), 110602(2020) CHINESE OPTICS LETTERS November 2020
1671-7694/2020/110602(6) 110602-1 © 2020 Chinese Optics Letters
资源评论
