Open Access
Add to BibSonomy
Abstract
Enabling laser white-lighting at a correlated color temperature (CCT) of 6500K with the use of only red/green/blue (RGB) tri-color laser diodes (LDs) is demonstrated, which can further perform wavelength division multiplexing (WDM) communication with a high-spectral-usage 16 QAM-OFDM data stream at 11.2 Gbps over 0.5 m. The sampling rate of encoded data is optimized to avoid the aliasing effect and to effectively amplify the signal with high on/off extinction and modulation depth. Proper oversampling can decrease the peak-to-average power ratio (PAPR) of the OFDM data and filter out unwanted noise. There are also six different diffusers used to diverge the white-light mixed by the RGB LD beam. By analyzing the color-casting transmittance, surface roughness, CCT uniformity, divergent angle of the diffuser, and the data transmission capacity, the frosted glass (FG2.8) diffuser with high transmittance diverges the white light with the divergent angle of ± 20° and supports the highest data rate of 14 Gbps over 0.5 m. To fit the day-light CCT, the blue LD power at an optimized bias current is further attenuated with a 0.6-optical density filter for reducing CCT from 100000K to 6500K; however, such an adjustment also degrades the SNR ratio to sacrifice the achievable data rate of the blue LD. The polycarbonate (PC1.5) diffuser with proper surface roughness diverged white-light exhibits the best CCT uniformity and a divergent angle of ± 30° but supports a data rate of only 6.4 Gbps over 0.5 m. The poly (methyl methacrylate) PMMA1.5 diffuser scatters the white light with the largest angle of ± 40°; however, the data rate also decreases to 4.8 Gbps over 0.5 m.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Recently, the solid-state white-lighting based optical wireless communication (OWC) [1] has attracted more attention than ever because of its well-known advantages including the high bandwidth and the high data security under immunity of electromagnetic interference and the high data security. These unique features make the white-lighting based light-fidelity (Li-Fi) suitable alternative to the wireless fidelity (Wi-Fi) for aircraft cabin and hospital become an environment. With the rapidly increased demand on the data transmission capacity at user end, visible optical carriers in the Li-Fi [2] module covering wavelength around 380-780 nm must to be simultaneously employed to efficiently support high-speed wavelength division multiplexing (WDM) data communication for avoiding the network traffic congestion. The conventional Li-Fi prototypes comprehensively use light emitting diodes (LEDs) in different colors for both white-lighting and data communication in hypermarkets, hospitals and even driverless cars [3–5]. As the illustrated application of Li-Fi in smart house shown in Fig. 1, the daylight lamps concurrently play the roles on multifunctional services including lighting, communication, position sensing and guiding, and incessant biomedical monitoring, etc. In the near future, the Li-Fi accompanied with street lamps and car lamps for pedestrians and vehicles will become normal scenario to play in human being society. In 2016, the eeNews EUROPE has foreseen that the global market of the highly secure VLC/Li-Fi module can achieve a revenue growth to USD 101.30 billion by 2024 [6]. In addition, the Business Wire also reported that the annual growth rate of the VLC market can increase by 94.55% to approach USD 10.81 billion in 2020 over a forecast period as the VLC will be regarded as supplementary for Wi-Fi, Bluetooth, and WiMAX [7].
Fig. 1 (a) The application of Li-Fi in smart house. (b) The picture and (c) experimental setup of 0.5-m RGB LDs WDM VLC system.
For white-light generation, there are two common approaches including the blue light-emitting diode stimulated phosphor (P-LED) and the mixed red/green/blue (R/G/B) LEDs. A P-LED based 1.5-m real-time VLC system with 37-Mbps data rate was accomplished by Yeh et al. [8]. Later on, a 100-Mbps non-return-to-zero (NRZ) VLC using a P-LED was implemented by Hoa et al. [9]. Previously, a 0.66-m bi-directional subcarrier-multiplexing (SCM) WDM visible-light communication (VLC) system with 225-Mbps downstream and 575-Mbps upstream based on P-LED and RGB LEDs was demonstrated by Wang’s group [10]. In addition, Jelena et al. have also used a single RGB LED luminary with the assistance of discrete multitone (DMT) modulation to demonstrate an 803-Mbps WDM VLC link [11]. By using a P-LED combining with the carrier-less amplitude and phase (CAP) modulation, a 1.1-Gbps white-light VLC system was preliminarily reported by Wu et al. [12]. Later on, the 3.4-Gbps RGB LEDs with DMT data format based WDM system was accomplished in 2012 [13]. A RGB-LED with CAP modulation based 4.5-Gbps WDM VLC system was also demonstrated by Wang et al [14]. The multivariable robust control of the correlated color temperature (CCT) and luminous/chromatic output for an R/G/B LEDs based lighting system was also proposed by Wang et al [15–17].
Alternatively, using laser diodes (LDs) to replace the LEDs could effectively enhance the transmission capacity of the proposed Li-Fi system because of its higher coherence, broader modulation bandwidth, larger output power and lower amplified spontaneous emission (ASE) noise background [18–24]. By employing red and green laser pointers, a 10m/500Mbps WDM VLC system was realized by Lin et al. [25]. At a free-space of 0.6 m, Chi et al. were the first group using 405- and 450-nm violet LD and blue laser diode (BLD) to excite a phosphor diffuser for demonstrating 4.4- and 5.2-Gbps Li-Fi system [26,27]. Subsequently, a tri-color R/G/B LDs based white-lighting communication beyond 8 Gbps was also reported by Wu et al. [28]. The stable violet/blue LD driven warm white light with CCT of 2700K and cool white light with CCT of 4400K was performed by Kristin et al. [29]. In view of previous works, the P-LED/P-LD based warm white-light with 2700-8000K VLC or RGB LDs based white-light with about 5000K have already been demonstrated [30,31]. In 2015, Janjua et al. mixed the beams of three individual R/G/B LDs to form white light [32]. By directly encoding the 16-QAM-ODFM upon each LD, the total data rates of the R/G/B LDs can be achieved to 4.4/4.0/4.0 Gbps with corresponding bit-error ratio (BER) of 3.3 × 10−3/1.4 × 10−3/2.8 × 10−3. In addition, Retamal et al. used the yellow phosphor encapsulated blue LD to obtain the mixed white light, and the BER can be improved to 2.8 × 10−5 with increasing the blue-light power when transferring the CCT of the white light from cool-day to neutral color [33]. Nevertheless, the realization of RGB 6500K-day-light based Li-Fi system is relatively difficult without the aid of yellow colored light source, which is left as an intriguing target to be investigated at current stage.
In this work, the R/G/B LDs mixed white-lighting source at 6500K is performed to implement the WDM VLC system with QAM-OFDM data transmission at 11.2 Gbps. To fit the demand of day-light at CCT of 6500K, the composition of R/G/B output powers need a precise adjustment; however, the sacrifice between demand of CCT and data rate are also observed and discussed. Under the narrow bandwidth modulation below 4 GHz, the sampling rate of the QAM-OFDM data individually encoded onto the LDs is adjusted to improve the allowable data rate.
2. Experimental setup
The white-lighting spot image and corresponding experimental setup of the R/G/B LD mixed white-light WDM Li-Fi are shown in Fig. 1, which consist of a 650-nm red LD (Mitsubishi, LPC-836), a 520-nm green LD (OSRAM, PLP 520) and 450-nm blue LD (OSRAM, PL 450B). By using individual thermo-electric coolers, all LDs were independently operated at the same room temperature of 25°C for stabilizing their output dynamics. Three plano-convex lenses (Thorlabs, LA1951-A) were employed to collimate the divergent tri-color laser beams into parallel lights, and three dichroic mirrors were subsequently used to combine the tri-color parallel laser beams for combining and mixing three beams into white light. In detail, the red parallel light was reflected by a mirror (Thorlabs, BBSQ1-E02) with a reflectivity of >99%, which is directed to combine with the green parallel light by using a dichroic mirror (Thorlabs, FF555-Di03). The red + green beams mixed to deliver a yellow parallel light, which was passing through another dichroic mirror (Thorlabs, FF495-Di03) to mix with the parallel blue laser beam for delivering the collimated white-light beam.
To form the divergent lighting spot without color aberration, six different kinds of diffusers were employed to diffuse/scatter the white-light laser beam. For testing the white-lighting performances after 0.5-m free-space transmission, the illuminance and efficiency, the commission internationale de l'éclairage (CIE) coordinate and the CCT of the R/G/B mixed white-light beam were measured by an illuminometer (TECPEL, DLM-530) and an integrating sphere assisted spectrometer (OKTEK, GL-2), respectively. Therein a neutral density filter with different optical density (O.D.) values is employed to avoid the saturation of the used spectrometer. For highly spectral usage data encoding, the electrical 16-QAM OFDM data with the same FFT size of 512 and a cyclic prefix (CP) length of 1/32 is generated. But versatile data streams with different subcarrier numbers were individually exported from an arbitrary waveform generator (AWG, Tektronix 70001A) at sampling rate varying ranged from 3 to 24 GS/s. Subsequently, a broadband pre-amplifier (Picosecond, 5865) with a power gain of 26.5 dB and a noise figure (NF) of 5.75 dB was employed to enlarge the peak-to-peak power of the electrical QAM-OFDM data.
After combining the amplified QAM-OFDM data with the DC bias current by using a bias tee (Mini-circuit, ZX85-12G-S + ) was directly modulated onto the RGB LDs for back-to-back and 0.5-m free-space transmissions. Note that all three LDs were encoded by carry 16-QAM OFDM data with different bandwidths to construct the WDM Li-Fi system. After free-space transmission, different bandpass filters were used to chromatically filter the corresponding red, green and blue light components from the mixed and delivered with light, and the filtered light was focused by using a plano-convex lens (Thorlabs, LA1131) prior to illumination the optical receiver. Afterwards, the optical 16-QAM OFDM data was received with a 1-GHz avalanche photodiode (APD, Hamamatsu, S12023-02) and amplified by a radio frequency (RF) low-pass amplifier (Mini-circuit, ZKL-1R5 + ) with a power gain of 40 dB, a bandwidth of 1.5 GHz and a NF of 3 dB, and then resampled by using a 100-GS/s digital serial analyzer (DSA, Tektronix, 71604C) with a resolution bandwidth of 16 GHz. Finally, a homemade MATLAB program was employed to analyze the error vector magnitude (EVM), signal-to-noise ratio (SNR) and BER of the received 16-QAM OFDM data after decoding.
3. Results and discussion
3.1 Characterization on the white-lighting property
The power-current-voltage (P-I-V) curve, the differential resistance (defined as dV/dI) and the differential quantum efficiency (defined as dP/dI) of the used RGB LDs were illustrated in Fig. 2(a)-2(d). Note that the threshold currents (Ith) for the R, G and B LDs are 78, 165 and 50 mA with related dP/dI slopes at I>Ith of 0.52, 0.19 and 0.49 W/A, respectively, as shown in Figs. 2(a) and 2(b). By differentiating the compliance voltage with the bias current shown in Fig. 2(c), the differential resistances of RGB LDs can be obtained and shown in Fig. 2(d). When the R, G and B LDs are operated beyond threshold, the corresponding differential resistances are 4.8, 3.2 and 9.6 Ω, which are much smaller than that of 50 Ω for other used RF components and instruments. This inevitably causes modulated signal reflection with induced reflection coefficients of (Γ) of −0.82, −0.88 and −0.68 for the R, G and B LDs, respectively. Moreover, the related voltage standing wave ratios (VSWRs) defined as VSWR = (1-Γ)/ (1 + Γ) are determined as 10.1, 15.7 and 5.25 to indicate the corresponding return losses (RL = −10logΓ2) of −1.73, −1.11 and −3.34 dB, respectively. To optimize the transmission, three 16-QAM OFDM data streams with bandwidths of 1-/0.3-/0.5-GHz were preset to individually encode the R/G/B LD biased at different currents. In principle, increasing the bias current can improve the modulation depth, enlarge the analog modulation, and suppress the relative intensity noise (RIN) of a laser diode. However, this operation often accompanies with the throughput power degradation at higher frequencies under larger bias, and the modulated data would be distorted owing to the power saturation at high bias current. With aforementioned constraints, the transmission of the RGB LDs can be optimized at 1.26, 1.3 and 1.8Ith to provide the lowest BERs of 7.0 × 10−4, 8.0 × 10−4 and 3.5 × 10−3, respectively, as compared to other cases. Such an optimized operation conditions of the RGB LDs are applicable for all diffusers used in the latter experiments.
Fig. 2 (a) The P-I, (b) dP/dI, (c) V-I and (d) dV/dI of RGB LDs. (e) The CIE coordinates and optical spectra of the RGB-LDs white lights diverged by the different diffusers with OD filter for attenuating the BLD power at different O.D. values. (f) The CIE 1931 color space and the CIE coordinates of the used RGB LDs.
To generate the cold white-light with CCT of 6500K which is located at CIE coordinates of (0.31, 0.33), the luminance of R/G/B components must be revised via detuning the output R/G/B LD powers without changing their biases. Before adjustment, the delivered white light exhibits an overly saturated CCT value owing to the enriched blue component which hazards human eyes. To decrease the CCT of generated white light for practical indoor lighting but maintain the available bandwidths of LDs for visible-light communication, an O.D. filter is used to attenuate the power of BLD beam. To diffuse the RGB mixed white-light for diverged beam expansion, six kinds of diffusers including a transparent frosted glass with a thickness of 2.8 mm (FG2.8), a both-side glossy polycarbonate with a thickness of 0.5 mm (PC0.5), a single-side frosted polycarbonate with a thickness of 1.0 mm (PC1.0), single-side tiger-skin frosted polycarbonate with a thickness of 1.5 mm (PC1.5), single-side frosted poly(methyl methacrylate) (PMMA) with thicknesses of 1.0 and 1.5 mm (PMMA1.0 and PMMA1.5) were employed for comparison. Increasing the O.D. value from 0 to 0.6, 0.9 and 1.1 attenuates the BLD beam power from 19.7 to 5.0, 2.5 and 1.6 mW, respectively, which makes the FG, PC and PMMA delivered white light with a daylight color at the same CCT of 6500K, as shown in Fig. 2(e). After attenuating the BLD power to obtain R:G:B power ratios of 3.1:2.0:1, 6.2:4.0:1 and 9.8:6.4:1, the CIE coordinates of the 6500K day-light diverged by the FG2.8, PC0.5, PC1.0, PC1.5, PMMA1.5 and PMMA2.0 diffusers are (0.30, 0.42), (0.30, 0.39), (0.30, 0.40), (0.30, 0.40), (0.30, 0.45) and (0.30, 0.43), respectively. Note that the CIE coordinates of six diffusers diverged white lights all deviate to the upper-left green-color space. With adapting the numerical calculation [32] for obtaining a white light at CCT of 6500K with its CIE coordinates exactly located at D65 (0.31, 0.33), the R/G/B LDs with CIE coordinates at R660 (0.73, 0.27), G520 (0.074, 0.83) and B450 (0.16, 0.018) are employed in the following procedure. Since the CIE coordinate of the white light at (0.31, 0.33) can be decomposed to R/G/B lights with certain weighting factors, such a white light CIE point should be located at the line connecting green light and purple light, and the purple light is located at the line connecting red light and blue light, as shown in Fig. 2(f). By finding the CIE coordinate of purple light, the power ratio of green and purple light components can be decided by calculating the distances from white light CIE to the lines of green and purple lights, and the power ratio is inversely proportional to the distances. Similarly, the power ratio of red light to blue light can also be calculated through the same procedure. In detail, the line connected red and blue lights can be expressed as yRB = 0.442x-0.053, and the line connected white and green lights can be expressed as yWG = −2.119x + 0.987. Hence, the CIE coordinate of the mixed purple light is located at (0.41, 0.13) that is the intersection point of two lines. By calculating the distance from purple light to red light and blue light, the red/blue power ratio of 11.87 and the green/purple power ratio of 2.59 can be observed [34]. Since the purple light is mixed by red and blue lights and the mixing ratio of green and purple lights is decided, the preliminary R:G:B power ratio is therefore determined as 4:11:1. Subsequently, the difference of human-eye sensitivity at the R/G/B wavelengths is further considered, and the R/G/B coefficients of 0.061/0.71/0.038 are implemented to correct the R:G:B power mixing ratio as 2.1:0.6:1 [35].
At a free-space distance of 0.5 m, the Fig. 3(upper) shows the uniformity of CCT value in lighting area for the RGB-LD mixed beam with different diffusers, which is measured by shifting the position of the illuminometer away from the collimated axis. Since the PC and the PMMA diffusers exhibit lower transmittances for the red laser diode (RLD) and green laser diode (GLD) beams than the FG, larger O.D. values of 0.9 and 1.1 are selected for attenuating the BLD beam to enable the generated white light a similar CCT of 6500K. Note that the vertical CCT values are higher than the horizontal CCT values for all diffusers because of the used LDs with unbalanced beam divergence on vertical and horizontal axes. For the FG2.8 diffuser diverged white light, the vertical and horizontal CCTs significantly decreases from 6500K to 5700K and from 6500K to 6300K at edge of the lighting area, respectively, as the edge roughness of the FG2.8 diffuser is larger than its central roughness due to the larger angle of incidence. In comparison, the CCT uniformities of the white light diverged by the PC0.5 and PC1.0 diffusers are similar on both vertical and horizontal axes owing to their similar spectral transmittances. For the PC1.5 diffuser, an ignorable CCT variance is observed to verify its omnidirectional uniformity, which for homogeneously diverges the RGB-LD beams. For the PMMA diffuser, the transmittance at blue wavelength is higher than that at red wavelength and the blue component is easier scattered than the red one, which makes the diverged white light show higher CCT at edge than that at center of the lighting area. For all diffusers, the scenario of the CCT adjustment is played by detuning the diffused RGB-LD mixing white light with the O.D. filters. The O.D. value of α = -log(T/100) characterizes the attenuation level of the O.D. filter, which means the transmitted power is 10-α of the original power with the larger α representing the stronger attenuation. In experiments, the employed O.D. filters with α of 0.6, 0.9 and 1.1 indicate the attenuation of 0.25, 0.13 and 0.08, respectively. The FG2.8 diffused RGB-LD white light requires only the O.D. filter with α = 0.6 to attenuate its blue component, which leaves the strongest power of the residual blue light to allow the highest transmission capacity for the proposed Li-Fi WDM system. In addition, the PC1.5 diffuser delivers white light with the most uniform CCT distribution to implement better illumination than others.
Fig. 3 (upper) The vertical and horizontal CCT uniformities and (down) the angle-dependent illuminations of the RGB LDs mixed white light diverged by the (a) FG2.8, (b) PC0.5, (c) PC1.0, (d) PC1.5, (e) PMMA1.0 and (f) PMMA1.5 diffusers.
In detail, the Fig. 3(upper) also shows the surface contour profiles of FG2.8, PC0.5, PC1.0, PC1.5, PMMA1.0 and PMMA1.5 diffusers taken by the atomic force microscopes (AFM), and the corresponding surface roughnesses (in root-mean square value) are determined as 172, 13, 55, 24, 51 and 51 nm, respectively. Note that the PC1.0, PMMA1.0 and PMMA1.5 diffusers have similar roughness for diverging the RGB LDs beam. The FG2.8 diffuser exhibits the roughest surface to make the diverged white light a significantly decreased CCT at the edge of the lighting area as confirmed in Fig. 3(upper). Among them, the PC0.5 diffuser reveals the flattest surface, which cannot effectively scatter the RGB LDs beam and some of the RGB LDs beam even reflected at surface. In application, the PC1.5 diffuser is therefore employed to make a compromise between the roughness and the flatness. Further, the angle-dependent illuminances of the RGB LDs mixed white light diverged by using different diffusers are shown in Fig. 3(down). Since the FG2.8 diffuser is a transparent glass with high transmittance, the diverged white light still exhibits high directivity with illuminance >300 lux with a divergent angle of only ± 20°, which would be the best candidate for data transmission. In comparison, the PC and PMMA diffusers with lower transmittance inevitably decrease the illuminance of the diverged white light to 20 lux. Both of the PC0.5 and PC1.5 diffusers show similar divergent angle of ± 30°, which are smaller than that of ± 40° for the white light diverged by PC1.0, PMMA1.0 and PMMA1.5 diffusers as flatter surfaces cannot equivalently scatter the RGB LDs beam owing to the wavelength dependent reflection.
3.2 Transmission performance of white-lighting Li-Fi WDM system
For VLC application, although the RGB LDs can carry high-bandwidth 16-QAM OFDM data, the receiving performance is limited by the low-bandwidth APD. Also, the diffuser used for beam divergence induces the throughput degradation to further suppress the receiving SNR of the QAM-OFDM data. Owing to such low data bandwidth, the current sampling rate as high as 24 GS/s is unnecessary for the proposed WDM Li-Fi system. In addition, the finite resolution of the AWG on the QAM-OFDM data amplitude inevitably induces additional noises. Therefore, different sampling rates of the FG2.8 diffuser diverged white light encoded 16-QAM OFDM data are expected for optimization. Note that there is no any OD filter used for attenuating the BLD beam power. After characterization, the received BERs and related constellation plots of the RGB LDs carried 1.4-, 0.4- and 0.6- GHz 1 6-QAM OFDM data at raw data rates of 5.6, 1.6 and 2.4 Gbps are plotted as a function of sampling rate ranged from 2 to 24 GS/s, as depicted in Fig. 4(a). For the green and blue LDs, the same optimized sampling rate of 3 GS/s is observed to indicate the lowest BERs of 2.6 × 10−4 and 2.1 × 10−5 when comparing with other sampling rates. Besides, the received BER of the RLD carried 5.6-Gbits/ data is improved from 5.3 × 10−3 to 4.1 × 10−4 by decreasing the sampling rate from 24 to 4 GS/s, whereas insufficient or over sampling below 3 GS/s or above 8GS/s leads to insufficient data amplitude, which cannot support the 1.4-GHz data bandwidth so as to degrade the received BER beyond 1.5 × 10−3. The proper oversampling effectively decreases the peak-to-average power ratio (PAPR) of the QAM-OFDM data and suppresses the background noise to improve the received SNR.
Fig. 4 (a) The subcarrier SNRs and related constellation plots of 1.4-/0.4-/0.6-GHz 16-QAM OFDM data carried by the R/G/B LDs at different sampling rates. (b) The PAPR of the 16-QAM OFDM data at the same bandwidth and different sampling rate.
Without inserting low-pass filter at receiver end, the modulated QAM OFDM data with different sampling rates are generated under the same fast Fourier Transform (FFT) size. The relationship between the subcarrier number, the bandwidth and the sampling rate is shown as
Under the same data bandwidth, the sampling rate and the subcarrier number are in inverse proportion each other. Figure 4(b) compares the PAPR of the received 16-QAM OFDM data with same bandwidth and different sampling rates. At high sampling rate, the PAPR is increased with decreasing subcarrier number which leads to decreased average power of the 16-QAM OFDM data, as the used AWG has a fixed peak amplitude output. At same peak amplitude, the high PAPR causes a low average SNR of the received 16-QAM OFDM data, hence the higher sampling rate leads to the higher PAPR and larger BER in our case.After receiving the RF spectra of 1.4-/0.4-/0.6-GHz wide 16-QAM OFDM data carried by the R/G/B LDs at different sampling rates are shown in Fig. 5(a), in which the frequency of the first subcarrier for all OFDM data is set at 0.14 GHz to avoid the low-frequency noises caused by the used electronic components. Note that decreasing the sampling rate respectively declines the throughput intensities by 9, 5 and 10 dB for the red, green and blue LD carried QAM-OFDM data, because of the suppressed PAPR accordingly. In addition, the throughput of the RLD carried data is larger than that carried by the GLD and BLD as the used APD receiver exhibits the highest responsivity at red wavelength. After optimizing the sampling rate of the extracted QAM-OFDM waveform from real-time oscilloscope, the deliverable RGB LDs carried 16-QAM OFDM data bandwidths can be increased to further enlarge the transmission capacity of the RGB-LD mixed white-lighting WDM Li-Fi system. To pass the forward error correction (FEC) criterion, the maximal allowable 16-QAM OFDM data bandwidths for encoding the red, green and blue LDs with their mixed white-light divergent by different diffusers are illustrated in Fig. 5(b), respectively. For the RLD, to approach the upper limitation on the transmission capacity, the maximal allowable 16-QAM OFDM data covering bandwidths of 1.5 GHz to give at raw data rates of 6 Gbps when using the FG2.8 diffuser. The other kinds of the diffusers inevitably degrade the allowable bandwidth and raw data rate to 0.4-1.0 GHz and 1.6-4.0 Gbps, respectively. As shown in Fig. 5(b), the data rate of RLD after divergent by the diffuser FG2.8 is important to show the highest capacity of data transmission among all cases, whereas the data rate obtained with the diffuser PC0.5 is the lowest one because of the flat surface induced color casting reflection of RGB LDs mixed beam.
Fig. 5 (a) The RF spectra of 1.4-, 0.4- and 0.6-GHz 16-QAM OFDM data carried by the RLD, GLD and BLD, respectively, at different sampling rates. (b) The subcarrier SNRs of the 16-QAM OFDM data carried by the RLD, GLD and BLD beams for mixing and diverging into white light by different diffusers.
The data rate in PC1.0, PMMA1.0 and PMMA1.5 cases reveal similar VLC performance due to their similar surface texture and roughness. Despite of the FG2.8 diffuser, the diffuser PC1.5 also shows appropriate roughness to uniformly scatter the white-light to the second high transmission data rate. For the GLD, its carried data rate with the FG2.8 and PC0.5 diffuser cases are respectively the highest of 3.6 Gbps and the lowest of 0.4 Gbps, as illustrated in Fig. 5(b), as elucidated by the same mentioned for the RLD case. When using the PC1.0 diffuser, the transmission performance deteriorates to 0.4 Gbps, which is smaller than the data rate obtainable with PMMA1.0 1.6 Gbps and PMMA1.5 1.3 Gbps owing to their transmittance of higher than that of the PC at wavelengths around 500~600 nm [28]. The subcarrier SNR performance of BLD divergent by different diffusers are shown in Fig. 5(b), Again the highest transmission capacities of BLD is observed as 1.1 GHz and 4.4 Gbps when using the FG2.8, whereas other cases can only provide 0.1-0.6 GHz and 0.4-2.4 Gbps, respectively. In such a RGB-LD mixed white-lighting system, the material and the roughness of the textured surface of the diffuser are mandatory for improving the transmission performance. After properly optimizing the sampling rate and by properly selecting the diffuser, the proposed RGB LD mixed white light can successfully function as a WDM Li-Fi system with allowable total data rate improved from 2.4 to 14 Gbps. Among three LDs, the RLD can provide the highest transmission capacity because it exhibits highest differential quantum efficiency and the used APD shows the highest responsivity of 35 A/W at such wavelength. The GLD and BLD only support lower transmission capacities due to the smaller lower responsivity of 12 and 5 A/W at green and blue wavelength regions. In particular, the GLD exhibits the worst modulation depth owing to its lowest P-I slope and largest return loss induced by impendence mismatch.
Figure 6 shows the UV-VIS transmittance spectra of different diffusers. For the FG2.8 diffuser which exhibits the highest transparency among all samples, its transmittance remains beyond 83% at around 630-650 nm, which exhibits the similar trend of transmittance versus wavelength with those of the PC0.5 and the PMMA1.0 samples. With the optical wavelength down-shifts from 660 nm to 450 nm, the FG 2.8 diffuser enlarges its transmittance from 83% to 92%. Such a transmittance enhancement is found for all samples used in this work. This elucidates the reason why the performance degradation on the BLD carried data is less than that on the RLD carried data after passing through these diffusers. Apparently, when the LD carried data transmits through the highly transparent FG2.8 diffuser, its intensity loss is the smallest one to preserve its data capacity and performance among all cases. In addition, a linear decrease on the transmittance of the PCx and PMMAx samples by 9 ± 3% are observed with the increment of every 50% on their increasing index. In comparison with other diffusers, the PC1.5 possesses the lowest transmittance at all wavelengths so as to provide the least allowable data rate of only 2.4 Gbps. The additional 15% decay on the overall transmittance would also result in the corresponding degradation on the SNR as well as the BER. From the abovementioned results, the transmittance and roughness of the diffuser may lead to a compromise between the data transmission and lighting chromatism of the white light carrier in the OWC link. As compared to other diffusers, the FG2.8 diffuser exhibits the highest transmittance and the smallest divergent angle for the white-light source mixed by the RGB LDs. Even though the spot size and coverage area of the synthesized white light are somewhat limited, the largest throughput power enables the FG2.8 diffused RGB-LD white light the highest receiving SNR at receiving end. Hence the highest transmission capacity with the best transmission performance is obtained by using the FG2.8 diffuser for the RGB-LD mixed white-light divergence.
3.3 Transmission performance of the white light at a CCT of 6500K
Although the maximal transmission capacity as high as 14 Gbps is observed for the RGB LDs mixed white-lighting WDM Li-Fi system, the frosted glass diverged white light still exhibits an overly saturated CCT value owing to the enriched blue component which hazards human eyes. Originally, the CIE coordinates of RGB LDs mixed white-light divergent by the FG2.8 diffuser is (0.23, 0.19), which located nearly the cold blue light region but away from the target with extremely high CCT. For delivering the 6500K white light with its CIE coordinates located at (0.31, 0.33), the BLD output power needs to be attenuated for applying to all diffuser cases. As PC and the PMMA diffusers seriously attenuate the BLD beam and make it too weak to carry the QAM-OFDM data, only the FG2.8 diffuser is considered in the following discussion, which enables the RGB-LD beam mixed and diverged into a qualified white-light spot, as shown in Fig. 7(a).
Fig. 7 (a) With the use of FG2.8 diffuser, the images of RGB-LD mixed and diverged at CCT = 6500K. (b) The received constellation related plots and spectrum of 1.1-, 1-, 0.8 and 0.4-GHz 16-QAM OFDM data carried by the RGB-LDs beam diverged with different diffusers in the WDM Wi-Fi system.
To check the lighting performance, the proposed WDM Li-Fi using R/G/B-LD mixed white-light system diverged by the diffuser FG2.8 illuminated on the paper is shown in Fig. 7(a). The white-light VLC output provides the cold white-light illumination with a transmission data rate of 11.2 Gbps after passing through 0.5-m distance in free space. The frosted glass FG2.8 enables the highest illuminance with the divergent angle of ± 20° provide a lighting area of 0.7 × 0.7 m2 square. In addition, the constellation plots and related RF spectrum of the BLD carried 16-QAM OFDM data in the proposed WDM Li-Fi system attenuated at different O.D. values are shown in Fig. 7(b). Increasing the O.D. values to 0.1, 0.3 and 0.6 respectively decreases the transmission capacity of the BLD to 4, 3.2 and 1.6 Gbps with corresponding EVMs of 16.6, 16.2 and 13.1%, respectively. All of the received data spectra show similar throughput-to-frequency slope and SNR, in which a noise peak located at around 1 GHz is observed because of the RIN of the BLD.
For the RGB LDs mixed white light diverged by the frosted glass, the data rate of the BLD after attenuating its power is shown in Fig. 8(a). Increases the O.D. value from 0 to 0.6 decreases the BLD power, and it also decreases the SNR as well as the BER such that the data rate of blue laser beam is dramatically reduced from 4.4 to 1.6 Gbps, which degrades the total transmission capacity of the WDM Li-Fi system to 11.2 Gbps. For detailed characterizations, the subcarrier SNRs of the 16-QAM OFDM data carried by the BLD component with different O.D. values are shown in Fig. 8(b). In particular, increasing the O.D. value from 0 to 0.6 slightly degrades the SNRs at low frequencies to worsen the maximal allowable 16-QAM OFDM data bandwidth of the blue laser beam from 1.1 to 0.4 GHz as the SNR significantly decays at beyond. In brief, the lighting and transmission performances of the R/G/B-LD beam mixed white light diverged by different diffusers are summarized in Table 1. Among these performances, the frosted glass FG2.8 diverged white light supports the largest transmission capacity of 14 Gbps within 0.5-m WDM Li-Fi distance, in which the achievable data rates of 6/3.6/4.4 Gbps are observed for the individual R/G/B LD beams. The performance of this work has already exceeded that of similar system with a total data rate of only 8 Gbps [28]. By using the OD filter with an O.D. value of 0.6 to attenuate the BLD power, the proposed white light approaches the lighting standard CCT of 6500K, CIE coordinate of (0.30, 0.42), illuminance of >300-lux illuminance and divergent angle of ± 20° at a cost of slightly decreased the transmission capacity (11.2 Gbps). In comparison, all the other diffusers (PC and PMMA) exhibit lower transmittances than the FG, their diverged white light beams exhibit weaker illuminances (~20 lux) to provide lower transmission data capacities. Even with the second best PC1.5 diffuser, the diverged white-light source with 6500K CCT shows the best uniformity with CIE coordinate (0.30, 0.40) and divergent angle ± 30° to support the total raw data rate of 6.4 Gbps. The PC0.5 diffuser with the flattest surface texture inevitably causes the largest color-casting reflection to degrade its transmission data capacity to 2 Gbps. Although the PC0.5 diffuser exhibits the worst transmission performance due to its lowest transmittance and largest color-casting reflection; however, the PC diffused RGB-LD white light can provide the larger divergent angle than the FG2.8 diffused beam.
Fig. 8 (a) The achievable bit rate and (b) the related subcarrier SNRs of the 16-QAM OFDM data carried by the BLD beam within the proposed white light diverged by the FG2.8 and attenuated at different O.D. values.
Table 1. The RGB LDs ratios, illuminations, materials, surface roughness, thickness, CCT variety, CIE coordinates and data rates of maximum and at CCT of 6500K.
The PMMA1.0 diffuser is used to scatter the R/G/B-LD mixed beam with the largest divergence of up to ± 40°, but its flattened surface degrades the achievable data rate down to 4.8 Gbps. The thicker PMMA1.5 diffuser reduces more data rate, as the transmittance and the roughness of the diffuser play important role on the transmission and lighting performances of the R/G/B-LD mixed white light. As the CCT of the mixed white-light source is strongly correlated with the relative power ratio of the RGB LDs, our work suggests to mutually detune the blue and red component powers for obtaining the CCT of 6500K from the outputs of different diffusers. Nevertheless, the wavelength dependent transmittances of diffusers deviate from one another although they have the same declination trend toward long wavelengths. In addition, the spectral responsivity of the used optical receiver is also decreased with shortening the illuminating wavelength. To compromise these effects for a minimal sacrifice on transmission capacity, the power adjustment of the BLD via different O.D. filters is appropriately considered for the white-light CCT optimization to 6500K. Owing to the least blue power attenuation after passing through the FG2.8 diffuser, its degradation on transmission capacity of the BLD is also the lowest among all diffusers. Two challenges of the RGB-LD mixed Li-Fi system for providing 6500K daylight are summarized as below. Due to the narrow linewidth of LD, the color-rendering index (CRI) of the RGB-LD mixed white light is still unqualified for demanded white-lighting standard as compared to the cool-white-light fluorescent lamp with CRI of 62 or >80 via the use of rare-earth phosphors. Insufficient CRI may not perfectly reveal the colors of illuminated objects under the RGB-LD white-lighting circumstances. Secondly, the CCT of the current RGB-LD synthesized white light is overly high without attenuating the excessive blue light, yet it seriously sacrifices the transmission capacity of data stream carried by the BLD, which leads to trade-off between lighting and communication.
4. Conclusion
The laser based white-lighting at CCT of 6500K with using the R/G/B LDs is demonstrated to perform the WDM Li-Fi communication up to 11.2 Gbps. For VLC application, the sampling rate of the encoded data is optimized to avoid the aliasing effect, to effectively amplify the data and to filter the unwanted noise. Note that the proper oversampling effectively decreases the PAPR of the QAM-OFDM data and suppresses the background noise to improve the received SNR. By analyzing the color-casting transmittance, surface roughness, CCT uniformity, divergent angle of the diffuser and the data transmission capacity of the individual LDs, the appropriate diffusers used for diverging the RGB LDs beam with optimized throughput of data transmission is obtained. Among them, the FG2.8 diffuser diverged white light supports the largest transmission capacity of 14 Gbps over 0.5 m. The standard cool-white-light CCT of 6500K is achieved after optimizing the power ratio of the RGB-LD mixed beam with a divergent angle of ± 20°, and a slightly decreased transmission capacity of 11.2 Gbps is obtained to enable the best Li-Fi performance among all diffusers. Such a white-light source with relatively small divergent angle is particularly suitable for high-speed and long-distance vehicle-to-vehicle communication. Other diffusers including PC and PMMA provide lower transmittance and data capacity than the FG does; however, their divergent angle can expand up to ± 40° for broadening the coverage of white-light spot. With diffusing by PC and PMMA, the RGB-LD mixed white light respectively allows 6.4-Gbps and 4.8-Gbps QAM-OFDM over 0.5 m at CCT of 6500K. With properly optimizing the sampling rate and selecting the scattered diffuser, the proposed RGB LD mixed white light can successfully function as a WDM Li-Fi system with its allowable total data rate greatly improved from 2.4 to 14 Gbps over 0.5 m.
Acknowledgments
We particularly thank Dr. C.-H. Cheng, Mr. H.-Y. Wang and Miss W.-C. Wang for the data analysis. We also acknowledge Mr. Y.-W. Hsueh for carrying out the experimental data.
Funding
Ministry of Science and Technology of Taiwan (MOST 104-2221-E-002-117-MY3, MOST 106-2221-E-002-152-MY3, and MOST 106-2218-E-005-001-).
References and links
1. L. Zeng, D. C. O’Brien, H. L. Minh, G. E. Faulkner, K. Lee, D. Jung, Y. Oh, and E. T. Won, “High data rate multiple input multiple output (MIMO) optical wireless communications using white LED lighting,” IEEE J. Sel. Areas Comm. 27(9), 1654–1662 (2009). [CrossRef]
2. D. Tsonev, S. Videv, and H. Haas, “Light fidelity (Li-Fi): towards all-optical networking,” Proc. SPIE 9007, 1–10 (2014).
3. A. Kumar, A. Mihovska, S. Kyriazakos, and R. Prasad, “Visible light communications (VLC) for ambient assisted living,” Wirel. Pers. Commun. 78(3), 1699–1717 (2014). [CrossRef]
4. A. M. Cailean, B. Cagneau, L. Chassagne, S. Topsu, Y. Alayli, and M. Dimian, “Visible light communications cooperative architecture for the intelligent transportation system,” In Proceedings of Communications and Vehicular Technology in the Benelux (SCVT) (Academic, 2013) pp. 1–5.
5. L. Grobe, A. Paraskevopoulos, J. Hilt, D. Schulz, F. Lassak, F. Hartlieb, C. Kottke, V. Jungnickel, and K.-D. Langer, “High-speed visible light communication systems,” IEEE Commun. Mag. 51(12), 60–66 (2013). [CrossRef]
6. J. Happich, “LED communications market worth $101.30 billion by 2024,” The eeNews EUROPE, http://www.eenewseurope.com/news/led-communications-market-worth-10130-billion-2024.
7. L. Wood, “Global Visible Light Communication Market - Expected to Reach $10.8 Billion by 2020 - Research and Markets,” Business Wire, https://www.businesswire.com/news/home/20171110005265/en/Global-Visible-Light-Communication-Market—.
8. C. H. Yeh, Y. L. Liu, and C. W. Chow, “Real-time white-light phosphor-LED visible light communication (VLC) with compact size,” Opt. Express 21(22), 26192–26197 (2013). [CrossRef] [PubMed]
9. H. Le Minh, D. O’Brien, G. Faulkner, L. Zeng, K. Lee, D. Jung, Y. Oh, and E. T. Won, “100-Mb/s NRZ visible light communications using a postequalized white LED,” IEEE Photonics Technol. Lett. 21(15), 1063–1065 (2009). [CrossRef]
10. Y. Wang, Y. Wang, N. Chi, J. Yu, and H. Shang, “Demonstration of 575-Mb/s downlink and 225-Mb/s uplink bi-directional SCM-WDM visible light communication using RGB LED and phosphor-based LED,” Opt. Express 21(1), 1203–1208 (2013). [CrossRef] [PubMed]
11. J. Vučić, C. Kottke, K. Habel, and K. D. Langer, “803 Mbit/s visible light WDM link based on DMT modulation of a single RGB LED luminary,” in Optical Fiber Communication Conference and Exposition (OFC/NFOEC),2011and the National Fiber Optic Engineers Conference, (Optical Society of America, 2011), paper OWB6. [CrossRef]
12. F. C. Wang, C. W. Tang, and B. J. Huang, “Multivariable robust control for a red–green–blue LED lighting system,” IEEE Trans. Power Electron. 25(2), 417–428 (2010). [CrossRef]
13. Y. Wang, X. Huang, L. Tao, J. Shi, and N. Chi, “4.5-Gb/s RGB-LED based WDM visible light communication system employing CAP modulation and RLS based adaptive equalization,” Opt. Express 23(10), 13626–13633 (2015). [CrossRef] [PubMed]
14. F. M. Wu, C. T. Lin, C. C. Wei, C. W. Chen, H. T. Huang, and C. H. Ho, “1.1-Gb/s white-LED-based visible light communication employing carrier-less amplitude and phase modulation,” IEEE Photonics Technol. Lett. 24(19), 1730–1732 (2012). [CrossRef]
15. G. Cossu, A. M. Khalid, P. Choudhury, R. Corsini, and E. Ciaramella, “3.4 Gbit/s visible optical wireless transmission based on RGB LED,” Opt. Express 20(26), B501–B506 (2012). [CrossRef] [PubMed]
16. X. Qu, S. C. Wong, and K. T. Chi, “Temperature measurement technique for stabilizing the light output of RGB LED lamps,” IEEE Trans. Instrum. Meas. 59(3), 661–670 (2010). [CrossRef]
17. M. Gilewski and A. Karpiuk, “An electronic control of light RGB LEDs,” Prz. Elektrotechniczny 84(8), 194–198 (2008).
18. T.-C. Wu, Y.-C. Chi, H.-Y. Wang, C.-T. Tsai, C.-H. Cheng, J.-K. Chang, L.-Y. Chen, W.-H. Cheng, and G.-R. Lin, “White-Lighting Communication with Lu3Al5O12:Ce3+/CaAlSiN3:Eu2+ Glass Covered 450-nm InGaN Laser Diode,” IEEE Photonics Technol. Lett. 36(9), 1634–1643 (2018).
19. H. M. Oubei, C. Li, K. H. Park, T. K. Ng, M. S. Alouini, and B. S. Ooi, “2.3 Gbit/s underwater wireless optical communications using directly modulated 520 nm laser diode,” Opt. Express 23(16), 20743–20748 (2015). [CrossRef] [PubMed]
20. Y.-F. Huang, Y.-C. Chi, H.-Y. Kao, C.-T. Tsai, H.-Y. Wang, H.-C. Kuo, S. Nakamura, D.-W. Huang, and G.-R. Lin, “Blue laser diode enforces free-space transmission up to 18 Gbps over 16 m,” Sci. Rep. 7(1), 1–8 (2017). [PubMed]
21. C. Lee, C. Zhang, M. Cantore, R. M. Farrell, S. H. Oh, T. Margalith, J. S. Speck, S. Nakamura, J. E. Bowers, and S. P. DenBaars, “4 Gbps direct modulation of 450 nm GaN laser for high-speed visible light communication,” Opt. Express 23(12), 16232–16237 (2015). [CrossRef] [PubMed]
22. Y.-C. Chi, D.-H. Hsieh, C.-T. Tsai, H.-Y. Chen, H.-C. Kuo, and G.-R. Lin, “450-nm GaN laser diode enables high-speed visible light communication with 9-Gbps QAM-OFDM,” Opt. Express 23(10), 13051–13059 (2015). [CrossRef] [PubMed]
23. S. Watson, M. Tan, S. P. Najda, P. Perlin, M. Leszczynski, G. Targowski, S. Grzanka, and A. E. Kelly, “Visible light communications using a directly modulated 422 nm GaN laser diode,” Opt. Lett. 38(19), 3792–3794 (2013). [CrossRef] [PubMed]
24. 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(18), 23302–23309 (2015). [CrossRef] [PubMed]
25. W. Y. Lin, C. Y. Chen, H. H. Lu, C. H. Chang, Y. P. Lin, H. C. Lin, and H. W. Wu, “10m/500 Mbps WDM visible light communication systems,” Opt. Express 20(9), 9919–9924 (2012). [CrossRef] [PubMed]
26. Y.-C. Chi, D.-H. Hsieh, C.-Y. Lin, H.-Y. Chen, C.-Y. Huang, J.-H. He, B. Ooi, S. P. DenBaars, S. Nakamura, H.-C. Kuo, and G.-R. Lin, “Phosphorous Diffuser Diverged Blue Laser Diode for Indoor Lighting and Communication,” Sci. Rep. 5(18690), 18690 (2015). [PubMed]
27. Y.-C. Chi, Y.-F. Huang, T.-C. Wu, C. T. Tsai, L. Y. Chen, H. C. Kuo, and G. R. Lin, “Violet laser diode enables lighting communication,” Sci. Rep. 7(1), 10469 (2017). [CrossRef] [PubMed]
28. T.-C. Wu, Y.-C. Chi, H.-Y. Wang, C.-T. Tsai, Y.-F. Huang, and G.-R. Lin, “Tricolor R/G/B Laser Diode Based Eye-Safe White Lighting Communication Beyond 8 Gbit/s,” Sci. Rep. 7(1), 11 (2017). [CrossRef] [PubMed]
29. K. A. Denault, M. Cantore, S. Nakamura, S. P. DenBaars, and R. Seshadri, “Efficient and stable laser-driven white lighting,” AIP Adv. 3(7), 072107 (2013). [CrossRef]
30. R. J. Xie, N. Hirosaki, N. Kimura, K. Sakuma, and M. Mitomo, “2-phosphor-converted white light-emitting diodes using oxynitride/nitride phosphors,” Appl. Phys. Lett. 90(19), 191101 (2007). [CrossRef]
31. S. Nizamoglu, G. Zengin, and H. V. Demir, “Color-converting combinations of nanocrystal emitters for warm-white light generation with high color rendering index,” Appl. Phys. Lett. 92(3), 031102 (2008). [CrossRef]
32. B. Janjua, H. M. Oubei, J. R. D. Retamal, T. K. Ng, C.-T. Tsai, H.-Y. Wang, Y.-C. Chi, H.-C. Kuo, G.-R. Lin, J.-H. He, and B. S. Ooi, “Going beyond 4 Gbps data rate by employing RGB laser diodes for visible light communication,” Opt. Express 23(14), 18746–18753 (2015). [CrossRef] [PubMed]
33. J. R. D. Retamal, H. M. Oubei, B. Janjua, Y.-C. Chi, H.-Y. Wang, C.-T. Tsai, T. K. Ng, D.-H. Hsieh, H.-C. Kuo, M.-S. Alouini, J.-H. He, G.-R. Lin, and B. S. Ooi, “4-Gbit/s visible light communication link based on 16-QAM OFDM transmission over remote phosphor-film converted white light by using blue laser diode,” Opt. Express 23(26), 33656–33666 (2015). [CrossRef] [PubMed]
34. L. B. Hooi, “Understand RGB LED mixing ratios to realize optimal color in signs and displays,” in LEDs Magazine (Academic, 2013).
35. E. F. Schubert, “Human eye sensitivity and photometric quantities,” in Light-emitting diodes (Academic, 2006).
