Performance Evaluation of FSO Communications under Sand-Dust Conditions

Cao, Minghua; Wang, Huiqin; Yao, Yu; Hou, Shanglin

doi:https://doi.org/10.1155/2019/2046896

International Journal of Antennas and Propagation

On this page

Abstract Introduction Conclusion Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2019 | Article ID 2046896 | https://doi.org/10.1155/2019/2046896

Performance Evaluation of FSO Communications under Sand-Dust Conditions

Minghua Cao,¹Huiqin Wang,¹Yu Yao,²and Shanglin Hou³

Academic Editor: Lorenzo Luini

Received27 Mar 2019

Accepted05 Oct 2019

Published31 Oct 2019

Abstract

Sand-dust weather conditions are considered the primary challenge to free-space optical (FSO) communications. It may cause severe attenuation that is malignant to FSO link performance. This study investigates the impact of sand-dust particles on a laser signal using the radiation propagation method and the small-angle approximation method. Numerical simulation shows that in sand-dust weather conditions, the multiple scattering effect is dominant and results in signal pulse delay and pulse broadening. Furthermore, the signal attenuation follows a negative exponential distribution to the laser wavelength. Superior performance can be achieved by employing a longer wavelength laser to reduce pulse delay and mutual interference.

1. Introduction

Free-space optical communication (FSO) is considered one of the most promising alternatives in metropolitan and local area networks as a viable solution to the access network last mile problem, in which a compromise between the available data rates and the cost is desirable [1, 2]. However, the performance of an FSO system is dependent on atmospheric conditions, particularly in the presence of aerosols (e.g., fog, smoke, sand-dust, and water vapour) which result in scattering and absorption effects [3–5]. Combined with aerosols, sand-dust particles may cause serious transmission interruption. Therefore, we focus on FSO link performance under sand-dust weather conditions to accurately predict the impact of sand-dust particles on FSO transmission.

Early researches on laser scattering primarily utilized Lambert–Beer’s law to analyse the single scattering of spherical particles [6–8]. However, it is insufficient to analyse only the single scattering scenario because the corresponding optical thickness increases with increasing transmission distance and atmospheric visibility. To resolve this issue, Islam proposed a simple model for the calculation of signal attenuation induced by sand-dust particles [9]. Wang investigated the influence of atmospheric visibility on pulse delay and pulse broadening through Mie theory and Monte Carlo method [10]. The spherical shape of the sand-dust particle is assumed in all these researches. However, the geometries of sand-dust particles cannot be classified as spheres, ellipsoids, cubes, or otherwise. Thus, the results may not be sufficiently accurate to describe the practical situation in FSO links.

On the problem of nonspherical particle scattering, Mackowski proposed a computational method to calculate the T-matrix of arbitrary-shaped particles by introducing moment solution to the discrete dipole approximation interaction equations [11]. Wang employed the finite element method to evaluate three types of boundary conditions and modelled the light propagation process in turbid media [12]. These researches provide novel perspectives to the problem of laser signal transmission under sand-dust weather conditions. Drawing inspiration from these researches, the radiation propagation method and the small-angle approximation method were utilized in this study to macroscopically analyse the characteristics of laser transmission in a sand-dust channel.

2. Theoretical Analysis

2.1. Laser Intensity

When a laser transmits in the scattering medium, its intensity can be expressed aswhere is the light intensity, e is the particle position, is the incident laser direction, is the scattering light direction from other particles, ρ is the particle density, and is the scattering phase function. is the extinction coefficient, where and represent the absorption coefficient and the scattering coefficient, respectively. Assume ɑ as the solid angle of the scattering light, as the solid angle of the direction, and as the radiation power of a unit volume along the direction. Consequently, equation (1) can be illustrated by Figure 1.

It is evident from the plot that the received signal consists of three main parts: (1) the attenuated signal in the direction which went through the particle scattering and absorption effects; (2) the signal from other directions, e.g., , which are superimposed on the direction; and (3) the signal emitted by the internal radiation source in the direction.

Assume the medium is completely absorptive and the photons follow the elastic scattering mode, and equation (1) can be simplified aswhere denotes the optical thickness. can be unfolded by the Legendre function aswhere θ and φ denote the zenith angle and azimuth angle of the incident light, respectively; θ′ and φ′ denote the zenith angle and azimuth angle of the emitting light, respectively; denotes the Legendre function; and denotes the Legendre constant.

In contrast, equation (2) can be expressed in Cartesian coordinates as

When it is assumed that μ = cos θ, , and j ≠ 0, . Furthermore, when the scattering phase function is assumed as and the scattering albedo as , equation (4) can be simplified as

However, the laser beam will gradually be deviated away from the optical axis due to the sand-dust particles’ multiple scattering effect. Therefore, the light intensity can be expanded by the sum of the scattered lights [13]:

If the scattered laser from other directions can be neglected, the light intensity can be derived by a recursive relation aswherewhere m is the scattering number which satisfies m ≥ 1.

Considering that the laser wavelength is much smaller than the average radius of the sand-dust particles, the sand-dust particles show a significant forward scattering selectivity in the light. Consequently, the laser power is concentrated mainly in a very small range along the propagation direction.

Substituting equations (7) and (8) into equation (6), it can then be expressed in the form of , where μ₀ denotes the initial value of cos θ before scattering:

According to the orthogonality of the Legendre polynomials, , where . Therefore, equation (9) can be rewritten as

According to small-angle approximation (where μ = μ₀ and ), equation (10) can be rewritten aswhere E₀ denotes the initial laser intensity.

Substituting equations (11) and (13) into equation (6), the light intensity of multiple scattering can be expressed as

In contrast, the laser intensity without scattering can be written aswhere δ is the Dirac function.

2.2. Laser Pulse Characteristics

The scattering of sand-dust particles leads to light intensity attenuation. It also results in laser pulse delay and pulse broadening. Suppose a laser beam is transmitting along the z-axis, and a sand-dust particle lies at the original point O. The laser propagation direction will be deflected to point P due to the scattering effect, as described in Figure 2. The average photon path length per unit time (denoted as dR/dt) can be expressed as

According to the classical electron scattering theory [14], the scattering projection angle α can be written aswhere θ is the scattering angle and . is the scattering phase function, which describes the laser power spatial distribution after scattering.

A revised two-term Henyey–Greenstein (TTHG) scattering phase function is proposed in [15]. The authors use two asymmetric factors to consider both forward scattering and backward scattering. Consequently, the function can be expressed aswhere and denote the asymmetric factor parameters, and the asymmetric factor , 0 ≤ λ ≤ 1, ≤ 0.

According to Koschmieder’s law [16], the relationship between atmospheric visibility () and the extinction coefficient (C_e) can be written aswhere λ is the laser wavelength, and q is a parameter related to atmospheric visibility, which is described by the following equations:

According to the particle forward scattering selectivity and the small-angle approximation theory, . Therefore, the x-direction component can be expressed as

Since the y-direction component of the XOZ plane is 0, the average laser scattering path length is

Integral to both sides of equation (23), the photon scattering path length R is

As a result, the pulse delay Δt can be expressed aswhere n denotes the complex refractive index of sand-dust particles, and c = 3 × 10⁸ m/s is the speed of light in a vacuum. It is evident from equation (25) that the laser pulse delay in sand-dust conditions is related to the parameters of transmission distance, optical thickness, scattering angle, scattering coefficient, and attenuation coefficient. Furthermore, the laser pulse delay is in direct proportion to the transmission distance and the sand-dust complex refractive index.

However, the pulse delay is in nanosecond order of magnitude, and the scattering process can be considered as a laser pulse throughout a linear time-invariant system. Consequently, the signal at the receiver end can be expressed aswhere denotes the laser pulse, which follows the Gaussian distribution and can be expressed aswhere T₀ is the pulse width. denotes the normalized pulse power, which can be expressed as [17]where is the unit pulse power of the received laser signal.

Substituting equations (27) and (28) into equation (26), the received signal is

The received signal is primarily related to the parameters of normalized pulse power, pulse delay, and pulse width.

Defining as a 3 dB pulse time extension width of ,

This indicates that 95.6% of the pulse power is concentrated in . Therefore, the amount of pulse broadening can be written as

It is evident from equation (31) that the amount of laser pulse broadening is affected primarily by the parameter of time delay and is related to the parameters of transmission distance, sand-dust particle refractive index, scattering albedo, optical thickness, and scattering angle.

2.3. Light Attenuation

Sand-dust concentration describes the quantity of sand-dust particles within a unit volume, which directly reflects the severity of the sand-dust weather condition [18]. The distribution of the sand-dust concentration with a certain particle size can be expressed as [19]where G = 2.65 × 10⁶ g/m³ denotes the sand-dust proportion; r denotes the radius of sand-dust particles; and denotes the sand-dust particle radius distribution, which normally follows the log-normal distribution [20].

Therefore, the probability density distribution function of the sand-dust particle radius can be written as

According to Koschmieder’s law, the relationship between atmospheric visibility and the attenuation coefficient C_e is

In terms of atmospheric visibility [21], can be expressed as

Additionally, the relationship between atmospheric visibility and antenna transmission height can be demonstrated as [22]where h is the vertical height, b = 1.25, and denotes horizontal visibility near the ground.

Substituting equations (36) and (37) into equation (32), the distribution of sand-dust particle concentration can be expressed as

However, sand-dust particles are randomly distributed in the air and can be considered as composite media composed of dry sand-dust particles and moisture. Its complex permittivity is determined by the dielectric constant of sand and moisture and varies with frequency. According to the Maxwell-Garnett equations, the complex permittivity of sand-dust particles takes the form [23]where ε_r and ε_i are the real and imaginary parts of , P denotes the volume fraction occupied by sand-dust particles in a unit volume, and and denote the complex permittivity of dry sand-dust particles and the relative dielectric constant of moisture, respectively. According to Wang et al. [24], the real parts and imaginary parts of are

According to the empirical formula proposed by Ellison [25], the real parts and imaginary parts of arewhere ; and α, λ_s, ε_1, and ε_∞ are four atmospheric temperature-dependent quantities. Consequently, the light attenuation in sand-dust weather conditions can be written as [26]:where, ε_r and ε_i are the real and imaginary parts of and λ is the laser wavelength.

Substituting equations (39)–(41) into equation (42), the light attenuation in sand-dust weather conditions can be derived as

3. Numerical Evaluation

Numerical simulations were carried out based on the preceding derivation. The simulation parameters were set as follows: the initial light intensity E₀ = 1, the signal pulse width T₀ = 60 ns, the temperature t = 20°C, and n = 1.55−0.005i, where the real part of n indicates the scattering capacity and the imaginary part of n indicates the absorption capacity.

3.1. Laser Intensity

According to Mie theory and equation (14), the laser intensity under single and multiple scattering scenarios are shown in Figures 3 and 4.

(a)

(b)

(a)

(b)

It is evident from Figure 3(a) that the laser intensity is directly proportional to the optical thickness and inversely proportional to the transmission distance in the single scattering scenario. Figure 3(b) shows that the laser intensity is inversely proportional to the scattering number and the optical thickness. The laser intensity follows exponential decay to the scattering number and the optical thickness.

Figure 4(a) illustrates the laser intensity versus the atmospheric optical thickness for different scattering number values. It can be observed that the maximum value is between 1 and 3. Furthermore, the laser intensity decreases rapidly with increasing scattering number values. This indicates that the multiple scattering will strongly attenuate the laser intensity in sand-dust weather conditions. In contrast, the laser intensity of multiple scattering is slightly larger than that of single scattering when the optical thickness is more than 5. This indicates that multiple scattering is dominant in the sand-dust channel when the optical thickness is high. The total received laser intensity versus the atmospheric optical thickness for different scattering number values is depicted in Figure 4(b). It is evident from the figure that the total received laser intensity increases with increasing scattering number values. The reason for this phenomenon is that the total received laser intensity is comprised of the single scattering light and the corresponding multiple scattering lights. Moreover, the total received laser intensity changes a little when the scattering number is more than three. Therefore, it can be inferred that the laser intensity of multiple scattering of more than three can be ignored.

3.2. Laser Pulse Characteristics

The received laser pulse waveform and the laser pulse characteristics under sand-dust weather conditions are shown in Figures 5–9 based on equations (15) and (29).

(a)

(b)

(a)

(b)

Figure 5 demonstrates the normalized laser pulse waveform at different transmission distances, where the scattering albedo is set as ω = 0.5, and the scattering coefficient is set as C_s = 7 × 10⁻⁴. The figure indicates that pulse width will broaden with increasing transmission distance. Table 1 shows the extended received pulse width with the different transmission distances, where the initial pulse width is set as 100 ns.

Figure 6 illustrates the effect of transmission distance and atmospheric visibility on laser pulse time delay under different laser wavelengths, where the laser wavelength becomes shorter from the bottom to the top. The pulse time delay is an increasing trend when atmospheric visibility decreases. Moreover, the short wavelength leads to greater pulse time delay.

The pulse time delay of different laser wavelengths at various transmission distances is shown in Figure 7(a), where the atmospheric visibility is set as 1 km. The pulse time delay grows incrementally over the transmission distance under different light wavelengths.

Table 2 shows the pulse time delay of different laser wavelengths, where transmission distance is 300 m. This illustrates that the pulse time delay gradually increases when using short-wavelength lasers as the carrier. Consequently, longer wavelength is conducive to reducing the pulse time delay and the mutual interference when transmitting signals over a short distance in sand-dust weather conditions. The influence of atmospheric visibility on pulse time delay is described in Figure 7(b), where the transmission distance is set as 1 km. It is evident from the plot that the decline of pulse time delay follows a negative exponential trend relative to the atmospheric visibility. Furthermore, the pulse time delay tends to be a wavelength-dependent constant when the atmospheric visibility is greater than 3.5 km.

Figure 8(a) demonstrates the curves of pulse time delay versus the transmission distance at different scattering albedos, where the scattering coefficient is set as 7 × 10⁻⁴. It is evident from the plot that the pulse time delay shows an exponential growth with increasing transmission distance. Table 3 shows the pulse time delay with the transmission distances, where ω = 0.6. Moreover, the pulse time delay increases with increasing scattering albedo. As a result, the multipath effect caused by scattering is the main factor behind the pulse time delay when the scattering effect is stronger than the absorption effect. Figure 8(b) shows the influence of the scattering coefficient on pulse time delay when ω = 0.7. It is evident from the figure that larger scattering coefficients result in longer pulse time delay. The reason for this phenomenon is that the larger scattering coefficient indicates higher sand-dust particle concentration in the atmosphere. Therefore, the probability of scattering is increased and leads to more serious multipath effect.

Figure 9 illustrates the relationship between pulse time delay and the asymmetry factor at different transmission distances when λ = 10.6 μm, = 1 km, and C_s = 7 × 10⁻⁴. As depicted in the plot, the pulse time delay decreases in a negative exponential trend with an increasing asymmetry factor. This is because forward scattering is predominant when the asymmetry factor gets close to 1.

3.3. Light Attenuation

The light signal attenuation characteristics under sand-dust weather conditions are shown in Figures 10–12 based on the result of equation (43).

(a)

(b)

(a)

(b)

Figure 10 depicts laser signal attenuation as a function of atmospheric visibility and humidity under three different laser wavelengths, where the laser wavelength becomes shorter from the bottom to the top. It is evident from the figure that the signal attenuation is inversely proportional to atmospheric visibility and directly proportional to atmospheric humidity. This phenomenon is particularly evident in the short laser wavelength scenario.

Figure 11(a) shows signal attenuation versus atmospheric visibility with different laser wavelengths in linear-log coordinates, where the atmospheric humidity is set as 10%. It is evident from the figure that the attenuation follows a negative exponential decay trend with increasing atmospheric visibility. The attenuation eventually tends to be a constant when the atmospheric visibility increases to 5 km. Furthermore, it is clear that the attenuation decreases with increasing laser wavelength. Figure 11(b) illustrates signal attenuation versus atmospheric humidity under different laser wavelengths. The attenuation increases with increasing atmospheric humidity.

Table 4 shows the signal attenuation of different laser wavelengths, where the atmospheric humidity is 5%, 15%, and 25%, respectively. This indicates that the signal attenuation decreases with increasing laser wavelength. As a result, we can conclude that longer laser wavelengths can help reduce signal attenuation when operating a laser communication system in sand-dust weather conditions.

Figure 12(a) demonstrates the relationship between sand-dust concentration and antenna transmission height under different levels of atmospheric visibility. It can be concluded from the figure that the sand-dust concentration shows a negative exponential decay trend to the antenna transmission height. Particularly, when the antenna transmission height increases beyond 2.5 km, the atmospheric sand-dust concentration reduces below 2.8 μg/m³ and gradually tends to be a constant. Figure 12(b) shows the influence of the antenna transmission height on signal attenuation with different laser wavelengths. It can be observed from the figure that the attenuation decreases with increasing antenna transmission height. Furthermore, it tends to be a constant value when the antenna transmission height exceeds 2.5 km. This is because the increased height leads to smaller atmospheric sand-dust concentration. Consequently, the effect of multiple scattering becomes less, which results in less attenuation.

4. Conclusion

We analysed the influential factors relevant to laser signal transmission in sand-dust weather conditions by utilizing the radiation propagation method and the small-angle approximation method. Simulation results were presented to evaluate the performance of laser signals transmitting through a sand-dust channel with scattering and absorption effects. Results indicate that sand-dust weather conditions cause significant light attenuation and pulse time delay. However, the results also indicate that longer laser wavelengths could potentially be utilized to improve the communication reliability of FSO systems in sand-dust weather conditions.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This work was supported in part by the NSFC Program (61875080, 61465007, and 61861026) and the Foundation of Gansu Education Department (2017A-011).

References

N. Kumar and H. Sohal, “Impact of various weather condition on the performance of free space optical communication system,” Journal of Optical Communications, vol. 35, no. 1, pp. 45–49, 2014.
View at: Publisher Site | Google Scholar
Z. Ghassemlooy, J. Perez, and E. Leitgeb, “On the performance of FSO communications links under sand-dust storm conditions,” in Proceedings of the IEEE International Conference on Telecommunications, pp. 53–58, Zagreb, Croatia, July 2013.
View at: Google Scholar
M. Ijaz, Z. Ghassemlooy, J. Perez, V. Brazda, and O. Fiser, “Enhancing the atmospheric visibility and fog attenuation using a controlled FSO channel,” IEEE Photonics Technology Letters, vol. 25, no. 13, pp. 1262–1265, 2013.
View at: Publisher Site | Google Scholar
M. A. Esmail, H. Fathallah, and M.-S. Alouini, “Outdoor FSO communications under fog: attenuation modeling and performance evaluation,” IEEE Photonics Journal, vol. 8, no. 4, pp. 1–22, 2016.
View at: Publisher Site | Google Scholar
M. Singh, “Performance analysis of WDM-FSO system under adverse weather conditions,” Photonic Network Communications, vol. 36, no. 4, pp. 1–10, 2018.
View at: Publisher Site | Google Scholar
H. C. V. D. Hulst and V. Twersky, Light Scattering by Small Particles, Dover Publications, Mineola, NY, USA, 1981.
I. I. Kim, B. McArthur, and E. Korevaar, “Comparison of laser beam propagation at 785 nm and 1550 nm in fog and haze for optical wireless communications,” in Proceedings of the SPIE Optical Wireless Communications III, vol. 4214, pp. 26–37, Boston, MA, USA, February 2001.
View at: Publisher Site | Google Scholar
L. K. Harada and D. H. Leslie, “Atmospheric transmission analysis for airborne laser application,” in Proceedings of the SPIE Laser Beam Propagation and Control, vol. 2120, pp. 56–73, Los Angeles, CA, USA, June 1994.
View at: Publisher Site | Google Scholar
M. R. Islam, Z. E. Omer Elshaikh, O. O. Khalifa, A. Z. Alam, S. Khan, and A. W. Naji, “Prediction of signal attenuation due to duststorms using MIE scattering,” IIUM Engineering Journal, vol. 11, no. 1, pp. 71–87, 2010.
View at: Publisher Site | Google Scholar
H. Q. Wang, Y. G. Wang, M. H. Cao, and X. Z. Ke, “Influence of atmospheric visibility on laser pulse delay and broadening in sand and dust weather,” Acta Optica Sinica, vol. 35, no. 7, Article ID 0701002, 2015.
View at: Publisher Site | Google Scholar
D. W. Mackowski, “Discrete dipole moment method for calculation of the T matrix for nonspherical particles,” Journal of the Optical Society of America A, vol. 19, no. 5, pp. 881–893, 2002.
View at: Publisher Site | Google Scholar
S. Hu, T. Gao, H. Li et al., “Light scattering computation model for nonspherical aerosol particles based on multi-resolution time-domain scheme: model development and validation,” Optics Express, vol. 25, no. 2, pp. 1463–1486, 2017.
View at: Publisher Site | Google Scholar
P. González-Rodríguez and A. D. Kim, “Light propagation in tissues with forward-peaked and large-angle scattering,” Applied Optics, vol. 47, no. 14, pp. 2599–2609, 2008.
View at: Publisher Site | Google Scholar
S. Nguyen-Kuok, “Erratum to: classical theory of the particle scattering,” in Theory of Low-Temperature Plasma Physics, Springer International Publishing, Berlin, Germany, 2017.
View at: Google Scholar
N. Pfeiffer and G. H. Chapman, “Successive order, multiple scattering of two-term Henyey-Greenstein phase functions,” Optics Express, vol. 16, no. 18, pp. 13637–13642, 2008.
View at: Publisher Site | Google Scholar
S. Shah, S. Mughal, and S. Memon, “Theoretical and empirical based extinction coefficients for fog attenuation in terms of visibility at 850 nm,” in Proceedings of the IEEE International Conference on Emerging Technologies, pp. 1–4, IEEE, Peshawar, Pakistan, December 2015.
View at: Publisher Site | Google Scholar
Y. M. Zhou, Q. Z. Liu, X. H. Zhang, and J. Sun, “An efficient method for simulating time-domain broadening of laser pulse propagating underwater,” Chinese Journal of Lasers, vol. 36, no. 1, pp. 143–147, 2009.
View at: Publisher Site | Google Scholar
Y. Yu, W. Huiqin, H. Qiu, P. Qingbin, and C. Minghua, “The air attenuation of laser transmission in sand-dust weather,” in Proceedings of the 16th International Conference on Optical Communications and Networks, pp. 1–3, IEEE, Wuzhen, China, August 2017.
View at: Publisher Site | Google Scholar
Q. S. Dong, “Physical characteristics of the sand-dust and sand-dust in different deserts of China,” Chinese Journal of Radio Science, vol. 12, no. 1, pp. 15–25, 1997.
View at: Google Scholar
R. D. Smirnov, A. Y. Pigarov, S. I. Krasheninnikov, M. Rosenberg, and D. A. Mendis, “Simulation of sand-dust statistical characteristics in tokamaks,” Contributions to Plasma Physics, vol. 48, no. 1–3, pp. 290–294, 2008.
View at: Publisher Site | Google Scholar
H. Li and X. Sang, “SNR and transmission error rate for remote laser communication system in real atmosphere channel,” Sensors and Actuators A: Physical, vol. 258, pp. 156–162, 2017.
View at: Publisher Site | Google Scholar
A. Ahmed, A. Ali, and M. Alhaider, “Airborne dust size analysis for tropospheric propagation of millimetric waves into dust storms,” IEEE Transactions on Geoscience and Remote Sensing, vol. GE-25, no. 5, pp. 593–599, 1987.
View at: Publisher Site | Google Scholar
A. H. Sihvola and J. A. Kong, “Effective permittivity of dielectric mixtures,” IEEE Transactions on Geoscience and Remote Sensing, vol. 26, no. 4, pp. 420–429, 1988.
View at: Publisher Site | Google Scholar
Z. Wang, Z. D. Fang, H. D. Ting, H. U. Tao, and W. L. Wen, “Calculation and simulation of sand and dust attenuation in microwave propagation,” High Power Laser and Particle Beams, vol. 17, no. 8, pp. 1259–1262, 2005.
View at: Google Scholar
W. J. Ellison, “Permittivity of pure water, at standard atmospheric pressure, over the frequency range 0–25 THz and the temperature range 0–100°C,” Journal of Physical and Chemical Reference Data, vol. 36, no. 1, pp. 1–18, 2007.
View at: Publisher Site | Google Scholar
E. A. A. Elsheikh, M. R. Islam, M. H. Habaebi, A. F. Ismail, and A. Zyoud, “Dust storm attenuation modeling based on measurements in Sudan,” IEEE Transactions on Antennas and Propagation, vol. 65, no. 8, pp. 4200–4208, 2017.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2019 Minghua Cao et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

821

Downloads

1133

Citations