Research Article Vol. 58, No.

16 / 1 June 2019 / Applied Optics 4553

Underwater wireless optical communication

system using a 16-QAM modulated 450-nm
laser diode based on an FPGA
JINGJING WANG,1,3 CHANGFENG TIAN,1 XINGHAI YANG,1 WEI SHI,1,* QIUNA NIU,1 AND T. AARON GULLIVER2
1
School of Information Science and Technology, Qingdao University of Science & Technology, Qingdao 266061, China
2
Department of Electrical and Computer Engineering, University of Victoria, Victoria, British Columbia V8P 5C2, Canada
3
e-mail: wangjingjing@qust.edu.cn
*Corresponding author: 03244@qust.edu.cn

Received 12 March 2019; revised 25 April 2019; accepted 5 May 2019; posted 8 May 2019 (Doc. ID 362297); published 31 May 2019

An underwater wireless optical communication (UWOC) system with a 450-nm laser diode and a 150-MHz PIN
photodiode is presented. The 16-ary quadrature amplitude modulation scheme is employed and realized on a
field-programmable gate array (FPGA) to efficiently provide a high data rate. Experimental results show that this
system can achieve data rates up to 50 Mbps over a distance of 3 m underwater with a bit error rate (BER) of
7.11 × 10−4 in artificial seawater at 35‰ salinity, where the attenuation coefficient is 0.481. This BER is below the
forward error correction target of 3.8 × 10−3 . Thus, this system can provide reliable high-speed communications
over short to moderate distance point-to-point UWOC links. © 2019 Optical Society of America
https://doi.org/10.1364/AO.58.004553

1. INTRODUCTION Table 1 summarizes the developed UWOC systems in terms

Oceans cover more than 70% of the Earth’s surface and have
attracted constant human exploration and development [1].
Recent studies into marine-resource exploitation and environ-
mental monitoring have accelerated the demand for reliable
underwater communication systems [2]. Therefore, improving
the data rates and reliability of these systems [3] is crucial to
ensuring adequate support for ocean-based applications.
Conventional underwater acoustic communication tech-
niques are most commonly used in ocean exploration.
However, the low bandwidth of their sound waves limits the
transmission data rates to tens of kilobits per second (Kbps),
and a significant correspondence delay makes it difficult to
meet the demand for reliable data-intensive applications [4].
The underwater radio frequency (RF) electromagnetic commu-
nication technique is advantageous in terms of its high band-
width, but its effective transmission distance is limited to
several meters by the high attenuation of RF waves in seawater
[5]. In contrast, underwater wireless optical communication
(UWOC) technology, which combines a high bandwidth
and data rates with low latency and power consumption in
a compact device, has emerged as a promising alternative for
satisfying increasing data-capacity requirements in underwater
wireless sensor networks (UWSNs) [6–10]. The existence of a
“light window” [11] in seawater allows a light transmission
distance greater than 100 m within a wavelength range of
450–550 nm. Consequently, blue-green light sources have
been developed for high-speed UWOC systems [12,13].
of their hardware schemes, light sources, modulation schemes,
achieved data rates, transmission distances, and bit error rate
(BER) [14–22].
It has been proved that light-emitting diodes (LEDs) and
laser diodes (LDs) are applicable to UWOC systems.
Although LEDs feature low cost and safe operation, their lower
modulation bandwidth (compared to LDs) restricts data rates
to tens of megabits per second (Mbps) in UWOC systems.
Thus, LDs have been extensively applied to UWOC
systems [23,24].
In addition, signal modulation and demodulation tech-
niques are the main components in these UWOC systems.
Studies of arbitrary waveform generator (AWG)-based and pro-
grammable pulse generator (PPG)-based UWOC systems indi-
cate that these can generate accurate and steady modulated
digital signals at gigabites per second (Gbps)-level data rates
while maintaining a low BER. However, as AWG and PPG
both depend on software signal processing via a PC, modula-
tion techniques are not implemented directly using hardware.
Therefore, most of the studied UWOC systems mentioned
above lack comprehensive design and implementation. In ad-
dition, the large size of experimental equipment also limits the
practicability of UWOC systems. In contrast, a field-program-
mable gate array (FPGA) can be used to realize modulation
directly with relatively low cost, smaller size, and higher prac-
ticability. Recent related research has mainly focused on simple
modulation schemes such as on-off keying (OOK), digital

1559-128X/19/164553-07 Journal © 2019 Optical Society of America

 4554 Vol. 58, No. 16 / 1 June 2019 / Applied Optics Research Article

Table 1. Comparison of the Characteristics and Performances of Various UWOC Systems

Hardware Scheme Light Source Modulation Scheme Data Rate Distance Bit Error Rate
PPG [14] 470-nm LED NRZ 10 Mbps 10 m (pool water)
AWG [15] 694-nm LD OOK 20 Mbps 30 cm (tap water) 3 × 10−4
AWG [16] 532-nm LD 16-QAM-OFDM 50.78 Mbps 2 m (tap water) 2.98 × 10−3
AWG [17] 520-nm LD 16-QAM-OFDM 1.118 Gbps 2 m (tap water) 2.98 × 10−3
AWG [18] 450-nm LD 16-QAM-OFDM 4.8 Gbps 5.4 m (tap water) 2.6 × 10−3
FPGA [19] 480-nm LED DPIM 1 Mbps 4 m (sea water)
FPGA [20] 470-nm LED RZ-OOK 5 Mbps 3 m (tap water) 1 × 10−6
FPGA [21] 488-nm LD PPM 10 Mbps 5 m (tap water) 7 × 10−8
FPGA [22] 448-nm LED NRZ-OOK 25 Mbps 5 m (tap water) 1 × 10−4
This work
FPGA 450-nm LD 16-QAM 50 Mbps 3 m (artificial seawater) 7.11 × 10−4

pulse interval modulation (DPIM), and pulse position A. 16-QAM Modulation

modulation (PPM), which limits the data capacity of UWOC The 16-QAM scheme is the main component of the proposed
systems [19–22]. Moreover, most experimental testing systems UWOC system in this paper as shown in Fig. 1. The upper and
involve fresh tap water instead of turbid water. lower parts of the figure are the modulation and demodulation
In this paper, a 50-Mbps underwater wireless laser commu- processes of the signal, respectively, and all modules are imple-
nication system with a 16-ary quadrature amplitude modula- mented using FPGA hardware. At the modulation unit, this
tion (16-QAM) scheme based on an FPGA is presented to paper utilizes a pseudo-random binary sequence (PRBS) as
efficiently provide a high data rate. The performance of this the data source of the UWOC system with a data rate of
system is tested under various levels of water turbidity. 50 Mbps, which is generated by a 15-bit cyclic shift register.
The sequence loop length is 215 –1, and its generation polyno-
mial is as follows:
2. OPERATION PRINCIPLES prbs  x 15  x 14  1: (3)
Visible light signals are subject to attenuation factors such as
absorption and scattering during transmission through under- The serial data stream is then converted into four parallel data
water channels, resulting in significant energy loss. These at- streams through serial-to-parallel conversion at a rate of
tenuation factors are represented by a
λ and b
λ, which 12.5 Mbaud/s. These data blocks are then mapped to 16-
can be superimposed as the c
λ of the underwater channel QAM symbols using a standard Gray code constellation,
while the three factors satisfy the following relationship: and the resulting symbols are divided into I and Q branches.
To mitigate the effects of inter-symbol interference (ISI), the
c
λ  a
λ  b
λ: (1) two branches symbols are up-converted and filtered to convert
to the baseband pulse shape. A finite impulse response filter
Table 2 shows the attenuation coefficient values for several dif- (FIR) is used for this purpose. To ensure sampling accuracy,
ferent types of water turbidity [25]. If we set the visible light we choose a sampling rate of 100 MSa/s in this study, equating
power before transmission through the underwater channel as to 8 times the 16-QAM symbol rate. The bandwidth of the
P i and the channel output light power as P t , in combination baseband pulse-shaping signals formula is as follows:
with the Beer–Lambert theorem, we can conclude that the
relationship between the two is as follows: B  R B
1  α: (4)
The roll-off coefficient α is set to 0.8, and thus the system
P t  P i exp−c
λd , (2)
bandwidth is 22.5 MHz, corresponding to a 50-Mbps trans-
where d is the light transmission distance. In this paper, the mission data rate, and bandwidth utilization is 2.22 bits/Hz.
attenuation coefficient c
λ of the test underwater channel will The pulse-shaping I and Q branch signals are then multiplied
be calculated based on this model, and the calculation results by the carrier signals with a frequency of 25 MHz, generated by
will be given in Section 4. a dual-channel direct digital synthesizer (DDS) at a sampling
rate of 100 MSa/s, and the multiplied results are summed to
obtain the 16-QAM signal.
Table 2. Attenuation Coefficients for Various Levels of Synchronization technology is an indispensable part of the
Water Turbidity demodulation unit. This paper implements two synchroniza-
tion techniques using an FPGA, including a carrier-synchroni-
Water Turbidity a [m−1 ] b [m−1 ] c [m−1 ] zation unit and an interpolation algorithm-based bit-
Pure seawater 0.0405 0.0025 0.043 synchronization (timing recovery) unit. These respectively re-
Clean ocean 0.114 0.037 0.151 cover the coherent carrier with the same frequency and same
Costal ocean 0.179 0.219 0.298 phase as the modulation module, and they provide optimal
Turbid harbor 0.266 1.824 2.19
sampling judgement times for the baseband pulse-shaping
 Research Article
Fig. 1. Block diagram describing the 16-QAM modulation module. DDS, direct digital synthesizer; LPF, low-pass filter; NCO, numerically
controlled oscillator.
signals to ensure that the data source from the transmitter is
accurately recovered. In the carrier-synchronization module,
two coherent carriers are generated by a DDS, whose frequency
and phase are adjusted in real time using phase-error detection
feedback through a phase detector and loop filter until they
equal those of the modulation module carrier. The clock rate
of the carrier synchronization unit is 100 MHz. After multipli-
cation with the modulated signal, the I and Q branch baseband
signals are recovered.
The bit-synchronization module first extracts and down-
converts the digital signal by the clock rate of 50 MHz. The
interpolated interval is generated by the numerically controlled
oscillator (NCO) and is input to the Farrow-structure-based
interpolation filter to calculate the two down-converted branch
signals and recover the optimal judgement time of the baseband
signals. The interpolated interval is adjusted through real-time
feedback, calculated using the Gardner-algorithm-based error
detector and a loop filter. The 16-QAM symbols are then
output at 12.5 Mbaud/s and recovered to the data source
by constellation demapping and parallel-to-serial conversion.
B. Bit Error Rate Module
This paper implements a real-time BER-monitoring module
based on an FPGA. The principle of this approach is initially
to create a demodulated data statistical counter and an error
counter. A reference PRBS is generated for comparison with
the demodulated PRBS. The error counter is triggered to run
when two sequences are not identical. An example simulation
Fig. 2. Bit–error-rate-monitoring module.
Vol. 58, No. 16 / 1 June 2019 / Applied Optics 4555
result is shown in Fig. 2, where “din” denotes the demodulated
data source, “flag” denotes the trigger of the reference PRBS,
“prbs” denotes the reference sequence, “error” denotes the error
bit counter, and “cntt” denotes the data bit counter, which has a
maximum data capacity of 1.34 × 108 bits.
3. EXPERIMENTAL SETUP
The real-time UWOC system proposed in this paper consists of
a transmitter unit, underwater optical channel, and receiver
unit. The generation of the 16-QAM signal, digital-to-analog
(DA)/analog-to-digital (AD) conversion, and demodulation of
the received signal are implemented using FPGA, while the
conversion of electrical signals to and from optical signals is
achieved using an optical transceiver. A block diagram for
the proposed UWOC link is given in Fig. 3. Table 3 summa-
rizes the system components and their specifications. The
transmitter unit is shown in Fig. 4(a). The source data is modu-
lated using the FPGA. After DAC, sampling at 100 MSa/s with
14-bits quantization accuracy, the amplitude of the resulting
16-QAM signal is adjusted via an amplifier and attenuator.
Then this signal is used to drive the TO-56 blue LD by a bias
tee to generate the 16-QAM optical signal. The modulation
bandwidth of the LD is over 100 MHz.
The receiver unit is shown in Fig. 4(b). A silicon-based
photodetector with an active area of 0.8 mm2 and a peak respon-
sivity of 0.44 A/W is used to detect the received optical signal.
After optical-electrical conversion, the analog signal is converted
 4556 Vol. 58, No. 16 / 1 June 2019 / Applied Optics Research Article

Fig. 3. Block diagram describing the proposed UWOC link. DAC, digital-to-analog conversion; ADC, analog-to-digital conversion; AMP,
amplifier; ATT, attenuator; Bias-T, bias tee; LD, laser diode; PIN, positive-intrinsic-negative; and MSO, mixed signal oscilloscope.

4. RESULTS AND DISCUSSION

Table 3. Summary of the Proposed UWOC System
Components A water tank of length 3 m is used to evaluate the performance
of the proposed UWOC system, with the transmitter and
Component Part No. Specifications receiver placed at either end as shown in Fig. 5.
FPGA Terasic DE2- Altera Cyclone IV Figure 6(a) shows the optical spectra measured under differ-
115 ent bias currents. There is a slight redshift in the spectra with
DAC HSMC DAC5672 250 MSPS (1–2 Vpp) increasing bias current, and the peak transmission wavelength is
Amplifier ZHL-6A-S+ 25 dB
Attenuator kT2.5-90/ 0–90 dB 450 nm corresponding to a bias current between 30–40 mA.
1S-2S The frequency response including the bias tee, laser diode, PIN
Bias current LDC205C 0-4 A (Bias-Current Range) photodiode, amplifier, and attenuator, measured under differ-
controller ent bias currents, is shown in Fig. 6(b). The −3-dB bandwidth
Bias tee LDM90 Ø5.6 mm, 600 MHz is 30.6 MHz for 20 mA, 130.9 MHz for 30 mA, and
(Modulation Frequency) 136.8 MHz for 40 mA. It can be seen that the −3-dB band-
Laser diode TO-56 >100 MHz (Modulation
(450-nm) Bandwidth) width of the system is greatly altered when the bias current is
PIN photodiode PDA10A-EC 150 MHz (3 dB Bandwidth) near 20 mA. This is because there is a threshold for the laser
MSO DSOX3104T 5 GSa/s driving current. When the current value is greater than the
ADC HSMC AD9254 150 MSPS (1–2 Vpp) threshold current, the −3-dB bandwidth is not significantly
affected by variations in the bias current.
Sea salt is then added to the fresh tap water to gradually
to a digital signal (with a sampling rate of 100 MSa/s) and cap- change the water turbidity. The transmitted optical power is
tured using a mixed-signal oscilloscope (MSO) to observe the determined for various bias currents as shown in Fig. 7(a).
signal waveform and spectrum. This digital signal is also Figure 7(b) depicts the received optical power for the different
demodulated using the FPGA. The resulting data is compared bias currents at 0–50‰ salinities of artificial seawater. The at-
with the reference PRBS to obtain the BER. tenuation of light passing through the wall of the water tank is

Fig. 4. (a) Transmitter components. (b) Receiver components.

 Research Article Vol. 58, No. 16 / 1 June 2019 / Applied Optics 4557

Fig. 5. UWOC system experimental setup.

Fig. 6. (a) Measured optical spectra for various bias currents. (b) Frequency response of the proposed UWOC system for various bias currents.

Fig. 7. (a) Curve of transmitted optical power versus bias current for LD. (b) Curve of received optical power versus bias current for different
salinity of artificial seawater.

measured separately. (The optical power of the receiver is to Eq. (2) as shown in Table 4. The following experimental
attenuated to 40%, like that of the transmitter.) After excluding results and analysis are all based on these calculated coefficients.
this factor, we calculate the attenuation coefficients for various Considering that the saturated optical power of a PIN pho-
levels of water turbidity when the distance d is fixed according todiode is 5 mW, we fix the bias current at the transmitter to
 4558 Vol. 58, No. 16 / 1 June 2019 / Applied Optics Research Article

Table 4. Calculated Attenuation Coefficients under 35 mA, which is greater than the threshold current. In this
Various Salinities of Artificial Seawater case, we test the bit error rate of the proposed system with fresh
Salinity (‰) 0 10 20 30 40 50
tap water for various gain values at the receiver. As shown in
Fig. 8, it can be seen that the optimal BER is achieved when the
Attenuation 0.012 0.088 0.342 0.432 0.541 0.878 gain value is set to 15 dB. The received signal is observed by a
Coefficient (m−1 )
MSO with an amplitude range of 900 mV–1.02 V as shown in
Fig. 9(a). We then interrupt the received signal and observe the
background noise introduced by the PIN photodiode, ampli-
fier, adjustable gain attenuator, and the RF line with the am-
plitude range of 50–87 mV, and thus the signal-to-noise ratio
can be estimated to be 22.99 dB.
The transmitted signal spectrum is given in Fig. 9(b). It
can be seen that the center frequency is 25 MHz, which equals
the carrier frequency, and the bandwidth of the transmitted
signal is 22.5 MHz, which matches the result provided in
Section 2.
The BER was tested for two different data rates of 4 Mbps
and 50 Mbps in fresh tap water. The obtained BERs
were 6.29 × 10−6 for 4 Mbps and 7.24 × 10−5 for 50 Mbps.
Furthermore, the BERs for various calculated attenuation
coefficients are given in Fig. 10.
The BER is 4.22 × 10−5 at a data rate of 4 Mbps, where the
Fig. 8. Bit error rate versus gain value at the receiver. c
λ is 0.481 (the salinity of artificial sea water is 35‰), a value
that increases with transmission data rate up to 50 Mbps, at

Fig. 9. (a) Transmitted and received waveforms captured by the mixed signal oscilloscope. (b) Spectrum of the transmitted 16-QAM signal.

Fig. 10. BER versus different attenuation coefficients.

 Research Article
which point the BER is 7.11 × 10−4 . These BERs are lower
than the forward error correction (FEC) criterion of 3.8 × 10−3 ,
with the corresponding constellations shown in Fig. 10.
5. CONCLUSION
An UWOC approach was implemented using an FPGA, blue
LD, and 150-MHz-bandwidth silicon-based photodetector. A
water tank of length 3 m was used to evaluate the performance
of the system experimentally. This was filled with artificial
seawater, and BERs were obtained for various attenuation co-
efficients of artificial seawater. For a data rate of 50 Mbps, the
BER corresponding to an attenuation coefficient of 0.481 (at
the average salinity of the world’s oceans), is 7.11 × 10−4 , which
is significantly lower than the FEC criterion. The system has a
small footprint and is shown to be robust to turbidity. Thus,
this compact and accurate system is suitable for underwater
short-to-medium-range high-speed communication applica-
tions. Future work will consider higher-order modulations such
as 64-QAM and 256-QAM to further improve the system
throughput.
Funding. National Natural Science Foundation of China
(NSFC) (61671261, U1806201); Key Research and Devel-
opment Program of Shandong Province (2016GGX101007);
University Science and Technology Planning Project of
Shandong Province (J17KA058, J17KB154); Natural Science
Foundation of Shandong Province (ZR2017BF036); Applied
Basic Research Projects of QingDao (17-1-1-4-jch).
REFERENCES
1. S. Levitus, J. I. Antonov, T. P. Boyer, and C. Stephens, “Warming of
the world ocean,” Science 287, 2225–2229 (2000).
2. C. Shen, Y. Guo, H. M. Oubei, T. K. Ng, G. Liu, K. H. Park, K. T. Ho,
M. S. Alouini, and B. S. Ooi, “20-meter underwater wireless optical
communication link with 1.5 Gbps data rate,” Opt. Express 24,
25502–25509 (2016).
3. X. Y. Liu, S. Y. Yi, X. L. Zhou, S. L. Zhang, Z. L. Fang, Z. J. Qiu, L. G.
Hu, C. X. Cong, L. R. Zheng, R. Liu, and P. F. Tian, “Laser-based
white-light source for high-speed underwater wireless optical commu-
nication and high-efficiency underwater solid-state lighting,” Opt.
Express 26, 19259–19274 (2018).
4. I. F. Akyildiz, D. Pompili, and T. Melodia, “Underwater acoustic sensor
networks: research challenges,” Ad Hoc Networks 3, 257–279
(2005).
5. Z. Feng and R. Allen, “Evaluation of the effects of the communication
cable on the dynamics of an underwater flight vehicle,” Ocean Eng.
31, 1019–1035 (2004).
6. Y. Q. Fu and Y. Z. Du, “Performance of heterodyne differential phase-
shift-keying underwater wireless optical communication systems in
gamma-gamma-distributed turbulence,” Appl. Opt. 57, 2057–2063
(2018).
7. Q. H. Tian, Z. Li, K. Hu, L. Zhu, X. L. Pan, Q. Zhang, Y. J. Wang, F.
Tian, X. L. Yin, and X. J. Xin, “Turbo-coded 16-ary OAM shift keying
FSO communication system combining the CNN-based adaptive
demodulator,” Opt. Express 26, 27849–27864 (2018).
Vol. 58, No. 16 / 1 June 2019 / Applied Optics 4559
8. L. W. Xu, J. J. Wang, Y. Liu, W. Shi, and T. Aaron Gulliver, “Outage
performance for IDF relaying mobile cooperative networks,” Mob.
Netw. Appl. 23, 1496–1501 (2018).
9. J. K. Zhang, S. L. Ding, H. L. Zhai, and A. H. Dang, “Theoretical and
experimental studies of polarization fluctuations over atmospheric
turbulent channels for wireless optical communication systems,”
Opt. Express 22, 32482–32488 (2014).
10. J. J. Wang, W. Shi, L. W. Xu, L. Y. Zhou, Q. N. Niu, and J. Liu, “Design
of optical-acoustic hybrid underwater wireless sensor network,”
J. Netw. Comput. Appl. 92, 59–67 (2017).
11. S. Q. Duntley, “Light in the sea,” J. Opt. Soc. Am. 53, 214–233
(1963).
12. H. H. Lu, C. Y. Li, H. H. Lin, W. S. Tsai, C. A. Chu, B. R. Chen, and
C. J. Wu, “An 8 m/9.6 Gbps underwater wireless optical communica-
tion system,” IEEE Photon. J. 8, 1–7 (2016).
13. T. C. Wu, Y. C. Chi, H. Y. Wang, C. T. Tsai, and G. R. Lin, “Blue laser
diode enables underwater communication at 12.4 Gbps,” Sci. Rep. 7,
40480 (2017).
14. N. Farr, A. D. Chave, L. Freitag, J. Preisig, S. N. White, D. Yoerger,
and F. Sonnichsen, “Optical modem technology for seafloor
observatories,” in OCEANS (IEEE, 2006), pp. 1–6.
15. A. Lin, W. Lu, J. Xu, H. Song, F. Qu, J. Han, X. Gu, and J. Leng,
“Underwater wireless optical communication using a directly
modulated semiconductor laser,” in OCEANS (IEEE, 2015), pp. 1–4.
16. J. Xu, C. Zhang, M. W. Kong, B. Sun, J. L. Wang, and B. Cong,
“On the characterization of a green-light directly-modulated diode-
pumped solid-state laser,” in International Conference on Optical
Communications and Networks (IEEE, 2017), pp. 1–3.
17. J. Xu, A. B. Lin, X. Y. Yu, Y. H. Song, M. W. Kong, F. Z. Qu, J. Han,
W. Jia, and N. Deng, “Underwater laser communication using an
OFDM-modulated 520-nm laser diode,” IEEE. Photon. Technol.
Lett. 28, 2133–2136 (2016).
18. H. M. Oubei, J. R. Duran, B. Janjua, H. Y. Wang, C. T. Tsai, Y. C. Chi,
T. K. Ng, H. C. Kuo, J. H. He, M. S. Alouini, G. R. Lin, and B. S. Ooi,
“4.8 Gbit/s 16-QAM-OFDM transmission based on compact 450-nm
laser for underwater wireless optical communication,” Opt. Express
23, 23302–23309 (2015).
19. M. Doniec, C. Detweiler, I. Vasilescu, M. Chitre, M. Hoffmann-Kuhnt,
and D. Rus, “AquaOptical: a lightweight device for high-rate
long-range underwater point-to-point communication,” Mar. Technol.
Soc. J. 44, 55–65 (2010).
20. J. A. Simpson, W. C. Cox, J. R. Krier, B. Cochenour, B. L. Hughes,
and J. F. Muth, “5 Mbps optical wireless communication with error cor-
rection coding for underwater sensor nodes,” in Oceans (IEEE, 2010),
pp. 1–4.
21. J. P. Liu, B. Zheng, L. F. Zhao, and Z. Y. Gong, “A design of under-
water wireless laser communication system based on PPM modulat-
ing method,” in Oceans (IEEE, 2016), pp. 1–6.
22. P. Wang, C. Li, B. Wang, and Z. Xu, “A cost-efficient real-time 25Mb/s
system for LED-UOWC: design, channel coding, FPGA implementa-
tion, and characterization,” J. Lightwave Technol. 36, 2627–2637
(2018).
23. M. W. Kong, Y. F. Chen, R. Sarwar, B. Sun, Z. W. Xu, J. Han, J. W.
Chen, H. W. Qin, and J. Xu, “Underwater wireless optical communi-
cation using an arrayed transmitter/receiver and optical superimposi-
tion-based PAM-4 signal,” Opt. Express 26, 3087–3097 (2018).
24. X. B. Sun, W. Q. Cai, O. Alkhazragi, E. N. Ooi, H. S. He, A. Chaaban,
C. Shen, H. M. Oubei, M. Z. M. Khan, T. K. Ng, M. S. Alouini, and B. S.
Ooi, “375-nm ultraviolet-laser based non-line-of sight underwater
optical communication,” Opt. Express 26, 12870–12877 (2018).
25. B. M. Cochenour, L. J. Mullen, and A. E. Laux, “Characterization of the
beam-spread function for underwater wireless optical communica-
tions links,” IEEE J. Ocean. Eng. 33, 513–521 (2018).