Non-line-of-sight methodology for high-speed wireless optical communication in highly turbid water

Journal Pre-proof Non-line-of-sight methodology for high-speed wireless optical communication in highly turbid water Xiaobin Sun, Meiwei Kong, Omar Alkhazragi, Chao Shen, Ee-Ning Ooi, Xinyu Zhang, Ulrich Buttner, Tien Khee Ng, Boon S. Ooi

PII: DOI: Reference:

S0030-4018(20)30018-3 https://doi.org/10.1016/j.optcom.2020.125264 OPTICS 125264

To appear in:

Optics Communications

Received date : 18 August 2019 Revised date : 31 December 2019 Accepted date : 5 January 2020 Please cite this article as: X. Sun, M. Kong, O. Alkhazragi et al., Non-line-of-sight methodology for high-speed wireless optical communication in highly turbid water, Optics Communications (2020), doi: https://doi.org/10.1016/j.optcom.2020.125264. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier B.V.

Journal Pre-proof *Manuscript-Redline Version Click here to view linked References

of

Non-line-of-sight methodology for high-speed wireless optical communication in highly turbid water XIAOBIN SUN,1 MEIWEI KONG,1 OMAR ALKHAZRAGI,1 CHAO SHEN,1 EE-NING OOI,1 XINYU ZHANG,1 ULRICH BUTTNER,2 TIEN KHEE NG,1 AND BOON S. OOI1,*

pro

1

King Abdullah University of Science and Technology (KAUST), Photonics Laboratory, Thuwal 21534, Saudi Arabia 2 King Abdullah University of Science and Technology (KAUST), Nanofabrication Core Lab, Thuwal 21534, Saudi Arabia *[email protected]

urn al P

re-

Abstract: Underwater wireless optical communication (UWOC) is an emerging technology for discovering, exploiting, and protecting various underwater resources. Due to the diverse water clarity conditions, light signals are expected to propagate in a fashion dominated by the lineof-sight (LOS) to non-line-of-sight (NLOS) regime. To fully benefit from the high capacity underwater internet that UWOC could offer, especially in the presence of turbid water, a system that obviates the stringent requirements for pointing, acquisition, and tracking (PAT) is required. Herein, we demonstrated a robust NLOS UWOC link fully relieving the requirement on PAT. Based on a system design consisting of an ultraviolet (UV) laser for enhanced light scattering and a high sensitivity photomultiplier tube (PMT), we established an NLOS link with a data rate of 85 Mbit/s and a transmission distance of 30 cm using on-off keying (OOK) in emulated highly turbid harbor water. Further, a data rate of 72 Mbit/s is still achieved when the alignment is totally lost, i.e., the pointing directions of the transmitter and receiver are parallel. A longer transmission distance up to 40 m is also envisaged. Our findings will pave the way for a practical, short-reach, NLOS UWOC link in realistic oceanic scenarios. Keywords: Non-line-of-sight, Underwater wireless optical communication, Ultraviolet, Water scattering

1. Introduction

Over 70% of the Earth’s surface is covered by water. To meet the demands of emerging underwater activities, such as tactical surveillance, environmental monitoring, marine archeology, and ocean rescue missions, the Internet of Underwater Things (IoUT) is expected to be developed in the near future [1]. Through integrating with existing underwater communication technologies, e.g., acoustic and radio frequency (RF) communication, underwater wireless optical communication (UWOC) is playing a vital complementary role in fulfilling the requirements of IoUT. Owing to its large modulation bandwidth (~MHz–GHz), low latency, high energy efficiency, and small-footprint transceivers, UWOC can provide data transmission over tens of meters at data rates of several Gbit/s [2–5]. Hence, UWOC will enable the creation of a worldwide network of smart interconnected underwater objects and digital connections throughout our oceans, streams, and lakes. The low water absorption coefficient in the blue-green window means that current UWOC studies mainly rely on light with wavelengths in the range 400–520 nm and a line-of-sight (LOS) configuration, which requires rigid alignment between the transmitter and the receiver [6]. Clear-water channels with low absorption and scattering are favorable for long-haul transmission using LOS UWOC. However, the real oceanic environment is more challenging in terms of maintaining the alignment. Oceanic turbulence [7,8], turbidity [9], and obstacles [7] in the water channel are reported to cause severe signal fading, and even complete signal loss,

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Journal Pre-proof

urn al P

re-

pro

of

in LOS UWOC links. To this end, a non-line-of-sight (NLOS) UWOC link was proposed to mitigate the aforementioned challenges. This can be configured either through light reflection from the water surface [10] or light scattering from minute particles in the water (e.g., plankton, particulates, dissolved organic matters, inorganics) [11]. As a result, the NLOS UWOC link can relive the strict pointing, acquisition, and tracking (PAT) requirements of LOS UWOC. Compared with surface-reflection NLOS, light-scattering-based NLOS circumvents the problem of signal fading caused by wave-induced variations in the water surface. In a lightscattering-based NLOS UWOC link, the photons emitted from the transmitter will be redirected multiple times by the minute particles in the water before being detected by the receiver. Therefore, it is feasible that the strict PAT requirements of LOS UWOC could be fully relaxed [6]. Herein, for the first time, we experimentally demonstrated a high-speed ultraviolet (UV) laser-based NLOS UWOC system in an emulated turbid water channel. For enhancing light scattering, a 377-nm UV laser was utilized for the NLOS system. Our previous work revealed that UV incurs lower path loss in NLOS UWOC compared to violet-blue light (405/450 nm) [12], and therefore the utilization of UV lasers [13] is further justified. Additionally, compared with other optical wavelengths, the solar radiation noise in the UV spectrum is the lowest on the Earth’s surface because of absorption through the ozone layer [14]. Figure 1 illustrates the complementary approaches that one could envisage for underwater internet, and calls for further research in the domain of NLOS communication.

Fig. 1. The schematic of UWOC links for underwater internet and broadcasting. The ship can deliver optical signals via a fiber cable to communicate with an autonomous underwater vehicle (AUV) through a LOS UWOC link. The submarine broadcasts light and communicates with the divers and AUVs. The communication between the submarines and the two divers shielded by the coral reef can be achieved through NLOS UWOC links. A diffuse-LOS UWOC link is established between the submarine and the diver who is covered within the broadcast area of the submarine. The submarine also communicates with one of the AUV through a LOS UWOC link.

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Despite the advantages of communication links in the UV region, the limitations in the transmitter and receiver technologies hindered the development in this regard. Practical UVB/UVC lasers are currently unavailable. For the receiver, high-performance photodiodes across the UV range with high sensitivity and response speed are crucial for communication applications. However, due to the low penetration depth of UV photons in the silicon layer, e.g., <20 nm for deep UV to UVA regime, the responsivity of commercial Si-based

Journal Pre-proof

2. Experimental details

urn al P

2.1 Transmitter

re-

pro

of

photodetectors usually exhibits less than 0.1 A/W for wavelengths below 400 nm, while further gradually decreasing towards the deep UV range [15]. Besides, severe inter-symbolinterference (ISI) caused by the multipath scattering process also impedes the development of UV-based NLOS UWOC. Being constrained with these issues, previous UV NLOS UWOC studies have relied on simulations. Monte Carlo simulations [16] and the Henyey-Greenstein (HG) phase function [17] were used to develop models describing the transmitted photons’ trajectory. The impulse response [11,18], bit-error-ratio (BER) performance [11,19], and the effects of channel geometry on path loss [20] have also been predicted based on theoretical simulations. Given the limitations mentioned above, in this work, we adopted a UVA laser (377 nm) as the optical transmitter. A photomultiplier tube (PMT) with high sensitivity (7 × 105 A/W) in the UV range was deployed to detect the weak received power. Aperture averaging mitigates the ISI through the use of an aspherical condenser lens with a large diameter (75 mm). Adopting the least-complex on-off keying (OOK) modulation technique, with a received power of 28 nW, we achieved a data rate of up to 85 Mbit/s through a highly-turbid water column. The BERs with respect to the path loss were also measured at various data rates. A longer transmission distance up to 40 m in a turbid NLOS UWOC link was further envisaged by considering the experimental results and a reported simulation study. We believe that our firstof-its-kind demonstration will open a new research paradigm complementing the LOS counterpart, especially in realizing stable and high-speed wireless communication in challenging water environments.

The transmitter used in this study is a TO-56 packaged UV diode laser (Nichia NDU4116). As shown in Fig. 2(a), the device has a center wavelength of ~377 nm, which is suitable for UVbased NLOS UWOC. However, multiple side modes are also present, which may introduce nonlinear noise to the signal modulation. The optical power-current-voltage (L-I-V) characteristics of the UV laser diode used in this study are shown in Fig. 2(b). The laser has a turn-on voltage of 3.078 V and a threshold current of 50 mA. The laser has a maximum operating current of 140 mA and a corresponding voltage of 6 V. Therefore, we limit the AC modulation voltage to less than 3 V. A bandwidth over 400 MHz is measured for this UV laser diode, this can be found in the supplementary information (Fig. S1).

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Fig. 2. (a) The normalized lasing spectrum of the UV laser diode. A peak wavelength of 377 nm and other lowerintensity modes exist at an operating current of 65.5 mA. (b) Optical power-current-voltage (L-I-V) plots for the UV laser diode. The turn-on voltage is 3.08 V and the threshold current is 50 mA.

Journal Pre-proof

2.2 Receiver

re-

urn al P

2.3 Water channel

pro

of

A PMT is an extremely sensitive light detector that provides a current output proportional to the intensity of the incident light. Therefore, though PMTs are typically used in photoncounting applications [21], they can also be used as detectors in optical wireless communication deploying intensity-modulation direct-detection (IM-DD) schemes, such as OOK or orthogonal frequency-division multiplexing (OFDM). Thus, a PMT is more preferred as the receiver in an NLOS UWOC link since it is able to provide high responsivity at the UV range through various amplification processes. The gain of the PMT will be tuned by the driving voltage. However, the amplification process obviously increases the response time of the PMT, which, in turn, limits the bandwidth of the PMT. A PMT (Hamamatsu R955) with a sensitivity of 7 × 105 A/W at 377 nm was used in this study. The input for the high-voltage controller was set to 2 V, corresponding to an allowable maximum incident power of 28 nW for the PMT, which is sufficient to ensure the feasibility of detecting intensity-modulated signals. The maximum incident power for the PMT under different inputs for the high voltage controller can be found in the supplementary information (Table S1). The PMT has a measured bandwidth of 37 MHz, and this can be found in the supplementary information (Fig. S2).

Fig. 3. (a) Particle size distribution of Maalox® solution. It is a Gaussian shape with a mean (μ) of 625 nm and a standard deviation (σ) of 128 nm. This leads to both Mie and Rayleigh scattering for 377-nm radiation. (b) Attenuation coefficients of the four types of emulated water for 377 nm radiation. Harbor II water has the largest attenuation of 2.19 m-1. The background shows scattering photos of a 377-nm beam propagating in the five water types.

In natural waters, particle sizes ranging from 0.01–1000 μm are of special interest to researchers concerned with the optical properties of seawater and light propagation in the sea [22]. In this investigation, four types of ocean water were considered: clear seawater, coastal water, and two types of harbor water, emulated by adding precise quantities of commercial antacid (Maalox®) to deionized water (DI). The particle and molecule size distribution in the Maalox solution was measured with Zetasizer and shown in Fig. 3(a). This is a Gaussian distribution with a peak particle size of 625 nm and a standard deviation of 128 nm. This distribution is within the interval range mentioned above and produces both Mie scattering (forward scattering) and Rayleigh scattering. Previous laboratory experiments have demonstrated that the Maalox solution presents excellent scattering features, similar to those of real ocean particles [23,24]. In addition, Laux et al. experimentally validated the use of Maalox as a scattering agent by

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Journal Pre-proof

pro

of

measuring and comparing its volume scattering function (VSF) against two different measurements [24], providing a baseline for determining the concentration of Maalox needed to emulate different ocean water types. We measured the attenuation coefficients at a wavelength of 377 nm in the four water types. As shown in Fig. 3(b), as the water turbidity increases, the attenuation coefficient increases. The DI water has the lowest attenuation coefficient (0.007 m-1), and the type II harbor water has the highest attenuation coefficient (2.19 m-1). The pure seawater, coastal water, and harbor I water have attenuation coefficients of 0.151, 0.398, and 1.1 m-1, respectively. The insets in Fig. 3(b) show the photos of the scattering effect when a 377-nm light beam propagates through the different water types. It is clear that the scattering is enhanced in the turbid water columns. For harbor II water, which has the highest attenuation coefficient, it is assumed that implementing a LOS UWOC link would be very difficult. However, as discussed above, UV-based scattering NLOS UWOC benefits from the effects of turbidity. In our case, the NLOS link is constructed in the harbor II water, which is the most turbid among the four types considered in this study.

urn al P

re-

2.4 Experimental setup

Fig. 4. The schematic of the experimental setup for UV laser-based NLOS UWOC. The setup consists of a ME 522A bit-error-rate tester, an 86100C digital communications analyzer, a UV laser diode, and a PMT. The communication link is a highly turbid water column emulating harbor II water, in which a certain amount of Maalox is mixed into the deionized water in a carefully prescribed fashion.

Figure 4 presents a schematic of the experimental setup for UV laser-based NLOS UWOC using a non-return-to-zero OOK (NRZ-OOK) modulation scheme. At the transmitter side, a UV laser and a UV-antireflection (AR)-coated collimation lens are mounted on a thermoelectric cooler (TEC) module (Thorlabs LD56F/M). The pattern generator (ME 522A Transmitter) is used to generate a pseudorandom binary sequence (PRBS) 210-1 pattern, and the data stream is fed into the UV laser. At the end of the transmitted beam in the water tank, a commercial beam dump is used to sink the incoming light photons to prevent the reflection from the tank wall, thus ensuring that all light received at the receiver side is from the scattering in the water. The photons scattered by various molecules in the water are then captured by the receiver. An ARcoated aspheric condenser lens (ACL 7560U-A) with a large diameter of 75 mm is installed in front of the receiver for aperture averaging, in order to reduce the multipath effects [25]. An adjustable iris with an optimal diameter of 0.8 cm is placed after the ACL for better signal

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Journal Pre-proof

urn al P

re-

pro

of

quality by allowing only the converged UV light propagating into the receiver. These photons then pass through a neutral-density filter (NDF) with an optical density of 2 (OD2, transmission = 1%) and are finally detected by the PMT, inducing the current signal. After passing through a transimpedance amplifier (TIA, Hamamatsu C6438-01) with a gain of 25 mV/μA, the current signal is amplified and converted to a voltage signal. The received signal is then sent back to the ME 522A receiver and a digital communication analyzer (Agilent Technology 86100C) for error ratio calculations and eye-diagram analysis, respectively. The signal-to-noise ratios (SNRs) for each data rate are also measured by the Agilent Technology 86100C. A synchronization clock is connected between the ME 522A transmitter and ME 522A receiver as well as between the ME 522A transmitter and an oscilloscope. The water tank used in this study has dimensions of 45 cm (length) × 30 cm (width). The distance between the transmitter and the receiver in this study, which is limited by the physical space, is 30 cm. To utilize the scattering of the whole water column, and to imitate the real ocean without any artificial boundaries, the transmitter has an azimuthal angle of 45° (T x = 45o) so that the beam can propagate from the right bottom corner to the top left corner of the water tank. Limited by the water tank size, the azimuthal angle of the receiver is regulated according to the received power and the NDF in front of it, and in this study, this angle is 23° (Rx = 23o).

Fig. 5. (a) The reflectivity for the black coating measured at the direct reflection position and diffuse reflection positions. The inset shows the schematic of the measurement based on a Newport 818-UV photodetector (PD). This represents the worst case scenario as the collimated laser beam is intentionally not directed into a beamdump. In actual measurements, the beam-dump serves to capture the collimated beam and eliminate a direct reflection. (b) Light guiding path characterization for the glass walls.

With the purpose of ensuring all possible paths contributed to the received signal except the NLOS scattering path in the experimental setup are effectively blocked, various methods and characterizations have been implemented to mitigate the reflective paths and also the possible light guiding paths. Firstly, reflective paths have been characterized and minimized. A separation board is installed between the UV laser diode and the PMT to prevent any scattering from air. Also, the transmitter and receiver are operated in a dark room and separated in two sealed black enclosures to prevent any reflection from other nearby objects. Furthermore, except for the incident glass wall of the water tank, the other inner walls of the water tank are covered with absorbing black coating to diminish the reflection from the glass walls. The reflectivity of the black coating is further characterized, as shown in Fig. 5 (a). The measurement at an angle of 0o means direct reflectivity, and the measurements at other angles are for the diffuse reflectivity. The results show that the implemented black coating absorber significantly reduces the reflectivity to a mere 0.0024% for direct reflection, and the diffused reflection for the black coating gets much weaker with increasing angle of measurement. By doing so, we excluded the possible reflection from the glass walls and the black coatings. Besides, the possible water-air interface reflection has also been characterized using the PMT by placing the black coating floating on the water surface. Using the same configuration

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Journal Pre-proof

pro

of

(Tx=Rx=90o, bias current for the laser of 65.5 mA) and Harbor II water type, we measured the received power of 13.68 nW with the black coating and 14 nW without the black coating. The small difference of 0.32 nW indicates that the water-air interface reflection contributes to a mere 2.3% of the received power. Additionally, we characterized the possible light guiding path, which is horizontally guided from the laser to the PMT by the glass in front of these devices. Note that some glasses are good light guiding mediums for some wavelengths [26]. As shown in Fig. 5(b), all the inner sides of the glass tank are covered with the black coatings, the transmitter and receiver sides are further enclosed in two separate enclosed boxes. By doing so, we limit the received power coming from the light guiding path through the glass wall. Under this configuration, a received power of 0.028 nW was measured using the PMT. Compared to the received power through the scattering link of 14 nW as mentioned above, the light guiding path from the glass wall contributes merely 0.2% to the total receiver power, which is negligible. In summary, the scattering paths, reflective paths, and light guiding paths have been clarified in this study, and only NLOS scattering paths are ensured to be active.

urn al P

re-

3. Results and discussion

Fig. 6. (a) Output current from the PMT responding to different laser bias currents. There is a linear response region when the laser bias current is in the range 40–90 mA. The maximum PMT output current is around 48 μA. (b) Measured BER of OOK modulation with respect to the bias current of the UV laser diode. The optimal bias current for the laser diode is 65.5 mA.

Figure 6(a) shows the PMT output current in response to the laser bias. A linear PMT response region appears when the laser bias current is in the range 40-90 mA. It gradually becomes saturated as the laser bias current increases above this range. Based on the PMT response curve, we set the bias current for the UV laser diode to 65.5 mA. An OOK signal of amplitude 3 V, fully utilizing the dynamic working voltage range of the laser diode, is set to obtain the maximum SNR, thus optimizing the channel performance. Additionally, the optimization of the bias current and the signal amplitude is further demonstrated by measuring the changes in BER with respect to the bias current at a data rate of 55 Mbit/s. As shown in Fig. 6(b), an optimized bias current of 65.5 mA provides the lowest BER for the system. According to the L-I-V curve of the laser diode, the corresponding output power from the UV laser diode under this bias current is 25 mW. The received radiation power can be derived from the corresponding PMT current:

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

preceived 

I output QE  S  Gain

(1)

Journal Pre-proof

where I output is the output current of the PMT after it detects the light signal, QE is the

urn al P

re-

pro

of

quantum efficiency of PMT cathode material, which is 25% according to the specification, S is the cathode sensitivity, which is 70 mA/W, and Gain is the gain from the PMT cathode to its anode during the amplification process, which is 107. The average received power for the PMT under the specific bias current of the UV laser is 14 nW, with a maximum received power of 28 nW (limited by the output current of the receiver system, which is 50 μA) when the UV laser reaches the maximum operating current.

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Fig. 7. (a) The frequency response of the system. The dashed line indicates the -3-dB bandwidth, (~31 MHz). (b) Measured BERs vs. data rates for the UV laser-based NLOS UWOC. The highest data rate achieved for this system is 85 Mbit/s with a measured BER of 3 × 10-3. (c) The measured eye-diagrams of OOK modulation under various data rates for the optical back-to-back (OBtB) transmission. (d) The measured eye-diagrams of OOK modulation under various data rates for the NLOS transmission.

Journal Pre-proof

urn al P

re-

pro

of

As a figure of merit, the small-signal frequency response of this NLOS UWOC system is shown in Fig. 7(a), taking into account both the transmitter (the UV laser diode) and the receiver (the PMT). A -3-dB bandwidth of ~31 MHz is measured. Given the fact that the UV laser and PMT have a bandwidth of over 400 MHz and 37 MHz, respectively, the modulation bandwidth of the system is, therefore, channel-limited. The BERs measured at various data rates are shown in Fig. 7(b). When the data rate increases, the BER gradually approaches the forward-error-correction (FEC) limit of 3.8×10-3. A BER of 3 × 10-3 is measured at a data rate of 85 Mbit/s. Fig. 7(c) shows the eye-diagrams for the received signal under various data rates. Open eyes are observed for the received signal with a BER below the FEC limit. As the BER nears the FEC limit, the eye-opening becomes smaller. The SNR decreases from 5.81 dB to 2.97 dB as the data rate increases from 30 Mbit/s to 80 Mbit/s. Note that the PMT is negatively biased, which makes the output negative. Therefore, the upper levels in the eye-diagrams represent logic 0, and the lower levels represent logic 1. Compared to the eyes obtained for optical back-to-back (OBtB) transmission, increasing accumulation of scattered data points can be seen at the logic 1. This nonlinear effect may be caused by the multiple modes in the laser, as previously mentioned [27]. Among those scattered photons that are collected by the PMT, it is highly possible that the path length differences between scattered photons result in temporal dispersion. In a sense, this temporal dispersion limits the link bandwidth, as mentioned above [28]. Further signal improvements can be achieved with low-noise elements or by using proper modulation schemes, e.g., RZ OOK, which offers better tolerance of nonlinear effects and noisy environments [29].

Fig. 8. The measured BER vs. received optical power at 50 Mbit/s, 52 Mbit/s, and 70 Mbit/s across the 30 cm transmitter-receiver distance.

Aiming to determine the maximum path loss for a specific data rate, we studied the BER performance as a function of the path loss over a 30-cm transmitter-receiver distance with data rates of 50 Mbit/s, 52 Mbit/s, and 70 Mbit/s. The received optical power is attenuated using a series of NDFs (Thorlabs NUK01) at the receiver side. The path loss is calculated from the received optical power:

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

P PL  10 log  t   Pr 

(2)

where PL is the path loss for the system; Pr and Pt are the received and transmitted optical power respectively. Figure 8 shows the measured BER with respect to the path loss. As shown in Fig. 8, the maximum path loss for maintaining a data rate with a BER below the FEC limit at 70 Mbit/s, 52 Mbit/s and 50 Mbit/s are 66.5 dB dB, 68.5 dB and 69.5 dB, respectively. With

Journal Pre-proof

urn al P

re-

pro

of

these path loss values, according to [20], NLOS UWOC links with transmission distances larger than 40-m can be envisaged. However, given the characteristics of range-dependent ISI, modulation schemes being capable of mitigating the data transmission errors brought from the increased ISI are required for future deployment [30]. To demonstrate that the requirement on PAT can be fully relieved in NLOS UWOC links, we further increase the azimuthal angles of both the transmitter and receiver. As shown in Fig 9(a), the azimuthal angle of both the UV laser and PMT are set at 90o (Tx = 90o, Rx = 90o), establishing a UWOC link with no alignment. By reducing the attenuation of the neutral density filter, the received optical power is controlled at the same range in this case to match that of the previous configuration (Tx = 45o, Rx = 23o). The achievable data rates using OOK are measured, as shown in Fig. 9(b). A maximum data rate of 72 Mbit/s with a BER of 3.7 × 10-3 is still obtained under this configuration. These results suggest that effective high-speed communication is still maintained even though the link totally lost its alignment.

Fig. 9. (a) The photo of the experimental setup when the communication link totally lost its alignment, i.e., Tx = 90o, Rx = 90o. (b) BER v.s. data rate when Tx = 90o, Rx = 90o.

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Fig. 10. (a) The measured SNRs for each data rate when Tx = 45o, Rx = 23o and Tx = 90o, Rx = 90o. (b) The measured impulse response for the case when Tx = 45o, Rx = 23o and Tx = 90o, Rx = 90o.

Journal Pre-proof

urn al P

re-

pro

of

On the other side, it should also be noted that the maximum achievable data rate of 72 Mbit/s is lower than the data rate of 85 Mbit/s achieved in the case of T x = 45o, Rx = 23o. The SNRs under this configuration for each data rate are measured and compared with those when T x = 45o, Rx = 23o, to investigate the possible reasons. As shown in Fig. 10 (a), with increasing of the data rate, the SNRs are decreasing. This can be explained from the limitation of the system bandwidth. Besides, due to the increased multiple scattering process, increased ISI during the signal propagation occurs when we increased the azimuthal angle of transmitter and receiver to 90o. This can also be approved by observing the impulse response of the systems. To investigate the temporal response of the link without being limited by the measured bandwidth of 31 MHz, we sent a pulse with a frequency of 10 MHz (100 ns/period). To simplify the analysis, we set the duty cycle at 50%. The total pulse width is of 52.5 ns takes into account the 50 ns trapezoidal pulse width, and 1.25 ns from half of the leading edge and 1.25 ns from half of the trailing edge. Under the two configurations, the received pulses with average mode are measured and shown in Fig. 10(b). As can be seen, the temporal dispersion for the configuration of Tx = 90o, Rx = 90o is slightly larger than that for the case of T x = 45o, Rx = 23o. Therefore, the SNRs for Tx = 90o, Rx = 90o are always smaller than those for T x = 45o, Rx = 23o. Hence, more proper modulation schemes to mitigate such ISI should be carefully investigated for the deployment of NLOS UWOC in the real environment. As discussed above, to further extend the transmitter-receiver distance and the data rate of the UV scattering-based NLOS UWOC systems, researches on both the transmitter and the receiver ends are critical. Single-mode high-power UV semiconductor lasers are needed, in particular, deep UV diode lasers. Ambient radiation-free at UVC region (100 - 280 nm) from the sun makes it feasible to reduce the noise significantly for an NLOS UWOC. The PMT used in the study provides high sensitivity at the price of the bandwidth in that it has a large detecting window (i.e., 8 × 24 mm). A UV photodetector simultaneously owning high sensitivity (i.e., > 105 A/W) and high speed (i.e., GHz) is demanding. By implementing with efficient modulation schemes, such developments of UV devices are supposed to help to build low-cost, high-speed, long-haul UV scattering-based NLOS UWOC. Various challenging underwater wireless optical communication links are becoming feasible, such as UWOC in turbid water, UWOC in an underwater environment full of obstacles, UWOC between an underwater node, and an autonomous underwater vehicle (AUV) with high motility. 4. Conclusions

We have demonstrated a UV laser-based NLOS UWOC system up to a data rate of 85 Mbit/s in a highly-turbid water channel based on an NRZ-OOK modulation scheme. An effective communication link with a data rate of 72 Mbit/s still holds when the channel lost its alignment completely, i.e., when the pointing direction of Tx is parallel to that of Rx. The proposed UWOC system employs a UV laser diode with an emission wavelength of 377 nm at an optimized bias condition, exhibiting a -3 dB bandwidth of 31 MHz. A longer transmission distance of up to 40 m is also estimated by comparing the BER vs. path loss studies with the reported theoretical literature. We believe that our first-of-its-kind NLOS demonstration will open a new research paradigm complementing the current LOS UWOC systems, especially in terms of realizing stable and high-speed wireless communication for challenging wireless underwater channel scenarios (i.e., turbid water column as well as underwater-object motility and mobility). Future UWOC system design should consider a UV-transmitter based NLOS modality in addition to the visible-light LOS modality for multi-scenario operability.

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Funding:

This work was supported by King Abdullah University of Science and Technology (KAUST) [BAS/1/1614-01-01, KCR/1/2081-01-01, and GEN/1/6607-01-01].

Journal Pre-proof

Acknowledgment

of

We gratefully acknowledge the personnel at KAUST for their relentless support of this work: Dr. Virginia Unkefer and Mr. Heno Hwang at the Publication Services and Researcher Support Department; Mr. Muhammad Q. Popalzai, Mr. Meshal M. Abdulkareem, and Mr. Michael Bayudan from the Central Workshop. Figure 1 was created by Heno Hwang, scientific illustrator at KAUST.

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13] [14]

re-

[2]

M.C. Domingo, An overview of the internet of underwater things, J. Netw. Comput. Appl. 35 (2012) 1879– 1890. doi:10.1016/J.JNCA.2012.07.012. C. Shen, Y. Guo, H.M. Oubei, T.K. Ng, G. Liu, K.-H. Park, K.-T. Ho, M.-S. Alouini, B.S. Ooi, 20-meter underwater wireless optical communication link with 1.5 Gbps data rate, Opt. Express. 24 (2016) 25502. doi:10.1364/OE.24.025502. X. Liu, S. Yi, R. Liu, L. Zheng, P. Tian, 34.5 m Underwater optical wireless communication with 2.70 Gbps data rate based on a green laser with NRZ-OOK modulation, in: 2017 14th China Int. Forum Solid State Light. Int. Forum Wide Bandgap Semicond. China (SSLChina IFWS), IEEE, 2017: pp. 60–61. doi:10.1109/IFWS.2017.8245975. Y. Chen, M. Kong, T. Ali, J. Wang, R. Sarwar, J. Han, C. Guo, B. Sun, N. Deng, J. Xu, 26 m/5.5 Gbps airwater optical wireless communication based on an OFDM-modulated 520-nm laser diode, Opt. Express. 25 (2017) 14760. doi:10.1364/OE.25.014760. C.-T. Tsai, D.-W. Huang, G.-R. Lin, Y.-C. Chi, Y.-F. Huang, Filtered Multicarrier OFDM Encoding on Blue Laser Diode for 14.8-Gbps Seawater Transmission, J. Light. Technol. Vol. 36, Issue 9, Pp. 17391745. 36 (2018) 1739–1745. https://www.osapublishing.org/jlt/abstract.cfm?uri=jlt-36-9-1739 (accessed May 7, 2019). H.M. Oubei, C. Shen, A. Kammoun, E. Zedini, K.-H. Park, X. Sun, G. Liu, C.H. Kang, T.K. Ng, M.-S. Alouini, B.S. Ooi, Light based underwater wireless communications, Jpn. J. Appl. Phys. 57 (2018) 08PA06. doi:10.7567/JJAP.57.08PA06. H.M. Oubei, R.T. ElAfandy, K.-H. Park, T.K. Ng, M.-S. Alouini, B.S. Ooi, Performance Evaluation of Underwater Wireless Optical Communications Links in the Presence of Different Air Bubble Populations, IEEE Photonics J. 9 (2017) 1–9. doi:10.1109/JPHOT.2017.2682198. H.M. Oubei, E. Zedini, R.T. ElAfandy, A. Kammoun, M. Abdallah, T.K. Ng, M. Hamdi, M.-S. Alouini, B.S. Ooi, Simple statistical channel model for weak temperature-induced turbulence in underwater wireless optical communication systems, Opt. Lett. 42 (2017) 2455. doi:10.1364/OL.42.002455. B. Cochenour, L. Mullen, A. Laux, Phase Coherent Digital Communications for Wireless Optical Links in Turbid Underwater Environments, in: Ocean. 2007, IEEE, 2007: pp. 1–5. doi:10.1109/OCEANS.2007.4449173. Shijian Tang, Yuhan Dong, Xuedan Zhang, On path loss of NLOS underwater wireless optical communication links, in: 2013 MTS/IEEE Ocean. - Bergen, IEEE, 2013: pp. 1–3. doi:10.1109/OCEANSBergen.2013.6608002. W. Liu, D. Zou, Z. Xu, J. Yu, Non-line-of-sight scattering channel modeling for underwater optical wireless communication, in: 2015 IEEE Int. Conf. Cyber Technol. Autom. Control. Intell. Syst., IEEE, 2015: pp. 1265–1268. doi:10.1109/CYBER.2015.7288125. X. Sun, W. Cai, O. Alkhazragi, E.-N. Ooi, H. He, A. Chaaban, C. Shen, H.M. Oubei, M.Z.M. Khan, T.K. Ng, M.-S. Alouini, B.S. Ooi, 375-nm ultraviolet-laser based non-line-of-sight underwater optical communication, Opt. Express. 26 (2018) 12870. doi:10.1364/OE.26.012870. R.C. Smith, K.S. Baker, Optical properties of the clearest natural waters (200–800 nm), Appl. Opt. 20 (1981) 177. doi:10.1364/AO.20.000177. D.M. Reilly, D.T. Moriarty, J.A. Maynard, Unique properties of solar blind ultraviolet communication systems for unattended ground-sensor networks, in: E.M. Carapezza (Ed.), International Society for Optics and Photonics, 2004: p. 244. doi:10.1117/12.582002. L. Shi, S. Nihtianov, Comparative Study of Silicon-Based Ultraviolet Photodetectors, IEEE Sens. J. 12 (2012) 2453–2459. doi:10.1109/JSEN.2012.2192103. C. Gabriel, M.-A. Khalighi, S. Bourennane, P. Léon, V. Rigaud, Monte-Carlo-Based Channel Characterization for Underwater Optical Communication Systems, J. Opt. Commun. Netw. 5 (2013) 1. doi:10.1364/JOCN.5.000001. C.D. Mobley, Light and water : radiative transfer in natural waters, Academic Press, 1994. J. V.K., A. Choudhary, F.M. Bui, P. Muthuchidambaranathan, Characterization of Channel Impulse Responses for NLOS Underwater Wireless Optical Communications, in: 2014 Fourth Int. Conf. Adv. Comput. Commun., IEEE, 2014: pp. 77–79. doi:10.1109/ICACC.2014.24. V.K. Jagadeesh, K. V Naveen, P. Muthuchidambaranathan, BER Performance of NLOS Underwater Wireless Optical Communication with Multiple Scattering, Int. J. Electron. Commun. Eng. 9 (2015) 563–

urn al P

[1]

pro

References

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

[15] [16]

[17] [18]

[19]

Journal Pre-proof

[23] [24]

[25] [26] [27] [28]

[29]

urn al P

[30]

of

[22]

pro

[21]

re-

[20]

566. https://waset.org/publications/10001375/ber-performance-of-nlos-underwater-wireless-opticalcommunication-with-multiple-scattering (accessed May 7, 2019). A. Choudhary, V.K. Jagadeesh, P. Muthuchidambaranathan, Pathloss analysis of NLOS underwater wireless optical communication channel, in: 2014 Int. Conf. Electron. Commun. Syst., IEEE, 2014: pp. 1– 4. doi:10.1109/ECS.2014.6892620. R. Foord, R. Jones, C.J. Oliver, E.R. Pike, The Use of Photomultiplier Tubes for Photon Counting, Appl. Opt. 8 (1969) 1975. doi:10.1364/AO.8.001975. M. Jonasz, G. (Georges) Fournier, Light scattering by particles in water : theoretical and experimental foundations, Elsevier/Academic Press, 2007. P. T, Volume Scattering Functions for Selected Ocean Waters, Scripps Inst. Oceanogr. (1972) 1–77. doi:10.5811/westjem.2013.7.18472. A. LAUX, R. BILLMERS, L. MULLEN, B. CONCANNON, J. DAVIS, J. PRENTICE, V. CONTARINO, The a, b, c s of oceanographic lidar predictions: a significant step toward closing the loop between theory and experiment, J. Mod. Opt. 49 (2002) 439–451. doi:10.1080/09500340110088498. D.L. Fried, Aperture Averaging of Scintillation, J. Opt. Soc. Am. 57 (1967) 169. doi:10.1364/JOSA.57.000169. J.S. Wang, E.M. Vogel, E. Snitzer, Tellurite glass: a new candidate for fiber devices, Opt. Mater. (Amst). 3 (1994) 187–203. doi:10.1016/0925-3467(94)90004-3. P.S. Donvalkar, A. Savchenkov, A. Matsko, Self-injection locked blue laser, J. Opt. 20 (2018) 045801. doi:10.1088/2040-8986/aaae4f. T.S. Rappaport, Wireless communications : principles and practice, Prentice Hall PTR, 2002. https://go.galegroup.com/ps/anonymous?id=GALE%7CA97115718&sid=googleScholar&v=2.1&it=r&lin kaccess=abs&issn=01926225&p=AONE&sw=w (accessed May 7, 2019). M.S.K. Faramarz E. Seraji, Eye-Diagram-Based Evaluation of RZ and NRZ Modulation Methods in a 10Gb/s Single-Channel and a 160-Gb/s WDM Optical Networks, Int. J. Opt. Appl. 7 (2017) 31–36. doi:DOI: 10.5923/j.optics.20170702.01. N. Kikuchi, Intersymbol Interference (ISI) Suppression Technique for Optical Binary and Multilevel Signal Generation, J. Light. Technol. Vol. 25, Issue 8, Pp. 2060-2068. 25 (2007) 2060–2068. https://www.osapublishing.org/jlt/abstract.cfm?uri=jlt-25-8-2060 (accessed August 4, 2019).

Jo

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Journal Pre-proof *Credit Author Statement

CRediT author statement XIAOBIN SUN: METHODOLOGY, VALIDATION, INVESTIGATION, WRITING-ORIGINAL WRITING-REVIEW & EDITING

of

MEIWEI KONG: INVESTIGATION, VALIDATION, WRITING-REVIEW & EDITING OMAR ALKHAZRAGI: INVESTIGATION, WRITING-REVIEW & EDITING

pro

CHAO SHEN: INVESTIGATION, WRITING-REVIEW & EDITING EE-NING OOI: INVESTIGATION XINYU ZHANG: INVESTIGATION

re-

ULRICH BUTTNER: RESOURCES

TIEN KHEE NG: PROJECT ADMINISTRATION, VALIDATION, WRITING-REVIEW & EDITING

Jo

urn al P

BOON S. OOI: CONCEPTUALIZATION, FUNDING ACQUISITION, SUPERVISION

DRAFT,