We experimentally demonstrate the implementation of 10 Gbps high-speed mid-infrared (MIR) free-space optical (FSO) communication, by means of our developed robust high-speed MIR transmitter and receiver modules. Such modules can enable frequency down- and up-conversion between 1550 nm and 3594 nm based on difference frequency generation (DFG) in MgO-doped periodically poled LiNbO3 (MgO: PPLN). The MIR transmitter generates 5.34 dBm power at 3594 nm for input powers of 33 dBm at 1550 nm and 37 dBm at 1083 nm. The MIR receiver regenerates −24.5 dBm power at 1550 nm for input powers of −1.2 dBm at 3594 nm and 36.7 dBm at 1083 nm. The eye diagram of regenerated 1550 nm signal is clear, and both for the on-off keying (OOK) and differential phase shift keying (DPSK) modulation, the power penalties compared with back to back (BTB) signals are lower than 3.5 dB measured at bit error ratio (BER) of 1E-6. According to our analyses, the system supports variable data rate, wavelength, and modulation format. Furthermore, the optical and electrical components are well integrated and fixed in MIR transmitter and receiver modules, which exhibit long-term stability and can be applied to field experiments.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Free space optical (FSO) communication has gained significant importance owing to its special features: high data rates, high security and unlicensed spectrum [1–3]. Versatile lasers based free-space communication has been realized [4–6] and the availability of FSO links has been proven by multiple demonstrations [7–9]. However, atmospheric attenuation and scintillation still play a significant role to the reliability of FSO links. Due to the favorable atmospheric properties, mid-infrared (MIR) region (3 μm ~5 µm) is a superior atmospheric transmission window compared with C-band region . Typically, the improvement of absorption loss for 3.6 μm compared to 1.5 μm in clean weather is ∼26% . The MIR is less affected by atmospheric turbulence, and has greater penetration through a bad weather condition such as fog. In addition, the spectral radiance of the main sources of background radiation in the atmosphere (Sun, Earth, Moon, etc.) has a pronounced minimum around MIR region, and thus the background noise will be minimized [12,13]. Yet, MIR communication has been hindered by lack of high-speed quantum-cascade lasers (QCLs), optical amplifiers  and quantum-well infrared photodetectors (QWIPs).
Up to now, many MIR communication studies have been carried out. Using QCLs to realize a MIR FSO link operating at 70 Mb/s with bit-error rate (BER) less than 1E-8 is reported by Soibel et al. . Direct modulation is typically limited to about 3 ~6 GHz for commercial QCLs due to the parasitic effects, and the modulation response bandwidth of QWIP is about 4.3 GHz in a 400 μm × 400 μm mesa geometry recently reported by Li et al. . Transmissions of analog audio/video signals and up to 2.5 Gb/s digital signals from an 8.1 µm wavelength QCL at temperature of 85 K is previously reported by Martini et al. [17,18]. Recently, A 3 Gb/s line transmission in four-level pulse amplitude modulation (PAM-4) configurations at 4.65 μm was achieved based on a QCL and a commercial infrared photovoltaic detector by Pang et al. . From the latest experiments mentioned above, utilizing QCL and QWIP to realize high speed (up to tens of Gbps) MIR communication is still a bottleneck due to the bandwidth limitation at the current stage, and another obvious drawback is that the modulation format is limited to only the intensity modulation such as on-off keying (OOK), pulse position modulation (PPM) and pulse amplitude modulation (PAM). However, compared to QCLs, other MIR laser sources such as difference-frequency generation (DFG) lasers , which can operate at very high modulation frequencies and are insensitive to modulation format [21–24]. Buchter et al. has demonstrated integrated optical transmitter and receiver modules at 3.8 µm by exploiting DFG in Ti-indiffused waveguides in periodically poled LiNbO3 (Ti:PPLN) and has successfully implemented a 2.488 Gb/s phase modulation transmission [25,26]. However, as the reported transmission data rate is all lower than 5Gbps, the performances can still be improved. In addition, the influences of the mutual conversion between the C-band and MIR on the intensity and phase of the signals are also not analyzed in .
In this work, we experimentally demonstrate a 10 Gbps differential phase shift keying (DPSK) transmission over 5-m long free-space link. To realize wavelength conversion between 1550 nm and 3594 nm based on DFG, we developed tunable MIR transmitter and receiver modules by exploiting frequency down- and up-conversion in MgO-doped periodically poled LiNbO3 (MgO: PPLN) [27,28]. The MIR transmitter generates 5.34 dBm power at 3594 nm and the MIR receiver regenerates −24.5 dBm power at 1550 nm. The calculated power penalty compared with back to back (BTB) signals are lower than 3.5 dB both for 10 Gbps OOK and DPSK modulation measured at BER of 1E-6. The influence of the DFG process on the wavelengths, data rates and modulation formats are also analyzed in detail. Furthermore, the well-integrated MIR transmitter and receiver modules exhibit long-term stability and can be applied to field experiments. To the best of our knowledge, this is the first demonstration of free-space DPSK transmissions at ten gigabit speed in MIR region.
2. Experiment setup
Efficient ultrafast all-optical wavelength converters and parametric amplifiers transparent to the data rate, wavelength and modulation-format are investigated in Fig. 1(a). Conversion between C-band and MIR region is realized through nonlinear DFG process in MgO: PPLN crystal. As shown in Fig. 1(a), the experimental configuration of MIR transmission system mainly consists of four parts: a C-band transmitter, a MIR transmitter, a MIR receiver and a C-band receiver.
2.1 C-band transmitter
At the C-band transmitter, continuous-wave light (around 5 kHz linewidth) generated from an external cavity laser (ECL, RIO0075-3-ITU-3) at 1550.12 nm is modulated in a 25 GHz Mach-Zehnder modulator (MZM, 25-PFA-PFA-LV) with a half-wave voltage of 4 V. An electrical data signal generated from an arbitrary waveform generator (AWG, Tektronix AWG70002) is amplified by a wideband electrical amplifier (EA, Photline DR-DG-10-MO-NRZ), and the amplified electrical signal is connected to the modulator radio frequency (RF) port through the SMA connector. The data pattern is a 215-1 pseudo-random bit sequence (PRBS) data stream. Simultaneously, an automatic bias control (ABC) board provides a precise voltage to maintain the modulator bias point at null or quadrature point for DPSK or OOK modulation with long-term stability. Thereafter, in order to improve optical signal to noise ratio (OSNR), a gaussian optical band pass filter (OBPF) with a bandwidth of 0.4 nm is added to narrow the signal spectrum. In addition, all components ensure the optical transmission with a linear polarization.
2.2 MIR transmitter
As shown in Fig. 1(a), the main functions of MIR transmitter are optical amplification and implementation of DFG process. The 1550 nm signal is amplified to 33 dBm (2 W) over two-stage amplification by using a polarization maintaining Er-doped fiber amplifier (PM-EDFA) and a double-cladding EDFA (DC-EDFA). The DC-EDFA is pumped by two laser diodes (LDs) at 976 nm with a total pump power of 16 W. An ytterbium-doped fiber laser (YDFL) with a wavelength of 1083 nm is used as a pump laser. The pump laser output linearly polarized light with a power of 12 dBm and then is split into two parts for the transmitter and receiver modules by a 50:50 beam splitter, respectively. Subsequently, similar to the amplification of 1550 nm signal, the pump laser is amplified to 37.3 dBm (5.4 W) through a polarization maintaining Yb-doped fiber amplifier (PM-YDFA) and a double-cladding YDFA (DC-YDFA). High power fiber isolators are placed between two amplifiers to suppress feedback-related output fluctuations.
The amplified 1550 nm signal and 1083 nm pump laser are combined by a high-power wavelength-division multiplexer (WDM) and emitted into the free-space through a fiber collimator Col1. The combined beam propagates through the center of a 50-mm 5% MgO-doped PPLN crystal (HC Photonics) by an achromatic lens embedded in the collimator Col1. A broadband half-wave plate (HWP) is employed to control the polarization states of two collimated beams for quasi-phase-matching (QPM) [18,19] condition. The PPLN crystal is coated at two surfaces with a transmittance of 99.5% at 1083, 1550, and 3594 nm. A followed 3.6-μm bandpass window F1 with a full width half maximum (FWHM) of 0.5-μm is used to filter the residual power at 1083 and 1550 nm. Both of Si-Ge lens L1 and L2 with a 50-mm focal length are used to collimate the MIR signal to a small divergence angle and accurately control the MIR beam propagation direction. Two golden mirrors M1 and M2 with a high reflection (> 99.5%) at 3.6-µm are used to establish a 5-m long FSO link between MIR transmitter and MIR receiver. Figure 1(b) shows the internal structure of the transmitter module. The optical and electrical components are fully integrated and assembled in the module. In addition, the functions of temperature control and device monitoring have been completely implemented.
2.3 MIR receiver
At the MIR receiver, the spatial MIR signal is focused into another same PPLN crystal by a Si-Ge lens L3 with a 100-mm focal length. The 1083 nm pump beam from fiber collimator Col2 is spatially combined and focused with the MIR signal by a flat-concave dielectric mirror M4 with a 50-mm radius of curvature. The HWP is used to adjust the polarization state of 1083 nm pump beam. The central position of signal and pump beams are carefully aligned to optimize the spatial overlap of the focal spots and also ensure that the overlap can be retained over long interaction lengths in the center of PPLN crystal for optimal conversion efficiency (CE). Then a followed dichroic mirror M5 and a BPF F2 are used to filter out the residual 1083 nm light in DFG light for improving the OSNR of 1550 nm signal. Figure 1(b) shows the internal structure of the receiver module.
2.4 C-band receiver
At the C-band receiver, a variable optical attenuator (VOA) is used to decrease optical power for performance evaluation. The 1550 nm signal is amplified to 20 dBm by an EDFA (Amonics, AEDFA-PA-35) which operates in automatic power control (APC) mode. The amplified signal is filtered through a programmable OBPF (Finisar, WaveShaper 4000S) with four output ports to primarily reduce the noise of amplified spontaneous emission (ASE) and wavelength conversion. The subsequent experiment settings in C-band receiver are designed to perform DPSK or OOK demodulation, eye-diagram observation and BER measurement. For OOK modulation, the OOK signal is directly detected by a 20 GHz photodiode (PD). For DPSK modulation, the DPSK signal is demodulated by a spatial structure Mach–Zehnder interferometer (MZI, Kylia MINT-1GHz) with a free spectral range (FSR) of 1 GHz and a balanced photodiode (BPD, Finisar BPDV2120R) with a 3-dB bandwidth of 43-GHz. At the end, the output serial electrical non-return to zero (NRZ) signal is measured by a 70 GHz digital sampling oscilloscope (DSO, Tektronix, TDS8200) and a bit error ratio test (BERT, Agilent N4961A) module with data rate up to 10 Gb/s.
3. Results and discussion
3.1 Signal generation in C-band
At the beginning of the experiment, we set the data rate to 10 Gbps and adopt DPSK modulation format. As shown in Fig. 1(a), optical spectrum and eye-diagram are measured at point A. Figure 2(a) illustrates the generated optical spectrum with an OSNR of 55.32 dB measured by a Fourier transform optical spectrum analyzer (FT-OSA, Thorlabs OSA 205) with a resolution of 7.5 GHz, Fig. 2(b) is eye-diagram of the DPSK signal measured by the DSO. The average power and center wavelength of the generated 10 Gb/s DPSK signal are measured as −2.43 dBm and 1550.118 nm, respectively.
3.2 C-band to MIR conversion
3.2.1 Generation of tunable MIR signal
In the MIR transmitter module, the 1550 nm optical signal requires two-stage EDFAs amplification, which the first stage reaches the order of hundred milliwatts for low noise and the second stage boosts to the order of several watts for high gain. As shown in Fig. 3(a), the output power of the DC-EDFA (DC-YDFA) increases approximately linearly with an average slope of 15.38% (38.03%) as a function of the laser diodes (LDs) power. In particular, maximum powers of 33 dBm (2 W) at 1550 nm and 37.4 dBm (5.4 W) at 1083 nm are obtained with the injected LD powers of 41.1 dBm (13.0 W) and 41.5 dBm (14.2 W), respectively. Both the bottom surfaces of two amplifiers are closely touched with the heat sink through thermal grease for large area heat dissipation. Figure 3(b) shows the output power stability of the two amplifiers in 35 minutes after the module is warmed up. The results show that power fluctuation of two amplified lasers are less than 1.2% with the temperature control at 0.5 °C accuracy.
After combining the two amplified lasers through the high power WDM, the combined beam is collimated and directly emitted to PPLN crystal for DFG process. DFG is a three-wave-mixing process where a strong pump wave () is combined with a (usually weak) signal wave () to generate a wavelength shifted idler wave (). The parametric process enables operation at arbitrarily low input signal powers, preserves signal phase information and reverses the sign of the signal phase. The conversion efficiency of DFG can be expressed as :Fig. 4(a). It is obvious that the MIR signal spectrum has a high OSNR of 35.66 dB. Figure 4(b) depicts the power fluctuation of the MIR signal in 35 minutes, which is about 0.3 dB power deviation after stabilization time of 5 minutes.
The developed MIR transmitter is a tunable laser source. Explicitly, we can vary the MIR signal wavelength by changing the wavelength of the injected C-band signal and the matching temperature, while keeping the pump laser wavelength fixed at 1083 nm. Figure 5 illustrates the simulation and experimental results of the DFG output power versus the operation temperature of the PPLN crystal. For simulation results, the DFG output power is calculated as Eq. (1), and the wavelength dependent phase mismatch in a QPM structure is expressed as [30,31]:Eqs. (1) and (2). For wavelengths of 1549.32nm, 1550.12 nm and 1551.72nm, the optimum temperatures for phase-matching are measured to be 53.62, 52.72 and 51.42 °C for a 30.41-μm grating period of the PPLN crystal. We take 1551.72 nm wavelength as an example, when the temperature increases from 42.6 to 51.4 °C, the measured normalized MIR power increases from 0.2 to 0.96, which indicates that the temperature is very crucial for optimization conversion efficiency of DFG in PPLN crystal. The deviation between the simulation results and the experimental results at three wavelengths are 2.41, 2.29 and 2.52 °C, which means that the simulation results are closely to the experimental results under the assumption of plane-wave interaction.
3.2.2 Compatible conversion with modulation format and data rate
To further investigate the impact of DFG parametric process on modulation format (such as OOK and DPSK) and data rate, we conduct more detailed experiments. Here, continuous wave (CW) light and modulated signal with three data rates of 1Gb/s, 6Gb/s and 10Gb/s are used for comparison and analysis. Figures 6(a)-6(b) depict the measured optical spectrum of CW light, OOK signal and DPSK signal at 1550 nm, respectively. Figures 6(c)-6(d) show the corresponding optical spectrum of MIR signal. Apparently, when the data rate becomes higher, both the spectrum profiles of OOK and DPSK signals at 1550 nm are broadened and the power decreases simultaneously. Figure 6(e) shows the output power of 1550 nm and 3594 nm signals as functions of the data rate. The results show that for OOK (DPSK) signal at 1550 nm, when the data rate increases from 1 Gbps to 10 Gbps, about 0.5 (0.6) dB power loss will arise. Correspondingly, about 3.5 (3.6) dB power loss will arise at 3594 nm. This means that the data rate increases by 10 times from 1 Gbps to 10 Gbps, while the conversion efficiency of OOK (DPSK) signal is dropped by 55.3% (56.3%). Because of one temperature point corresponds to an optimum frequency point, while high speed (wideband) signals consist of many frequency points, which leads the conversion efficiency to drop in a single temperature.
Figures 6(c)-6(d) show the spectral width of MIR signals with OOK and DPSK modulation at different data rate. Here, we define a factor k ( = Δν/Δν0) to characterize the spectrum broadening ratio. Δν0 is the 3-dB spectral width of 1 Gbps signal. The calculated factor k values are plotted versus data rate, as shown in Fig. 6(f). For OOK (DPSK) modulation format, from 1 Gbps to 10 Gbps, the spectral width is broadened by 3.31 (3.37) times at 1550 nm, while it is about 1.71 (1.64) times at 3594 nm. Because of the ASE noise induced from the 1550 nm and 1083 nm amplifiers fill a part of the optical spectrum, the spectral width of the 3594 nm signals with data rates from 1 to 10 Gbps are very similar.
3.3 MIR to C-band conversion
3.3.1 Compatible conversion with modulation format and data rate
The MIR signal is propagated through a 5-m FSO link through two golden mirrors. In order to regenerate the 1550 nm signal, the MIR signal and the pump signal at 1083 nm are coaxially coupled into PPLN crystal by precisely adjusting the position of the mirror M4 as shown in Fig. 1(a). Under the optimum polarization state of pump beam and matched temperature to approach the phase matching condition of the PPLN crystal, DFG light is obtained for the injected powers of −1.2 dBm at 3594 nm and 36.7 dBm at 1083 nm. The generated DFG light contains the 1550 nm signal and residual 1083 nm light, which are transmitted and reflected through the mirror M5 with a 99.5% transmittance at 1550 nm and 99.5% reflectivity at 1083 nm. After passing through the filter F2, the measured actual power and wavelength at 1550 nm is −24.5 dBm and 1550.121 nm. The regenerated 1550 nm signal is then coupled into a single mode fiber (SMF) with a coupling loss of 6.8 dB and is observed by the FT-OSA, as shown in Fig. 7(a). It can be seen that the profile of the regenerated optical spectrum is similar to the spectrum of the transmitted signal shown in Fig. 2(a).
Referring to the analysis method in subsection 3.2.1, Fig. 7(b) shows the output power of regenerated 1550 nm signal as a function of the data rate. The powers of regenerated OOK and DPSK signals with data rates from 1 to 10 Gbps are both distributed in the range of −19 dBm to −24.5 dBm. Correspondingly, the power coupled into the fiber ranges from −25.8 dBm to −31.3 dBm. Figures 7(c)-7(d) depict the optical spectrum of the regenerated CW light, OOK signals and DPSK signals respectively, whose spectrum profile is stretched at the bottom compared with those shown in Figs. 6(a)-6(b), but the signals still remain the sidelobes on both sides. This indicates that in nonlinear parametric process, the 1083 nm pump and the MIR signal interact to produce most of the 1550 nm signal power and remain the characteristics of the original 1550 nm signal due to quasi-phase-matching, while the 1083 nm and the ASE noise and nonlinear noise interact very weakly due to phase randomness of noise.
3.3.2 Conversion efficiency comparison
After establishing C-band–MIR–C-band FSO link, we compare the conversion efficiency (CE) of MIR transmitter and MIR receiver as functions of data rate shown in Fig. 8. From 1 to 10 Gbps, for OOK and DPSK modulation, the measured CE of MIR transmitter and MIR receiver are distributed in the range of 0.17% to 0.38% and 0.45% to 0.70%, respectively. The conversion efficiency of MIR receiver is higher than transmitter conforming to the nonlinear wavelength conversion formula [32–34]. The formula shows that conversion power is proportional to the square of the wavelength, so the conversion efficiency of long-wavelength to short-wavelength is higher than the opposite conversion.
3.4 Signal demodulation
At the C-band receiver, after low noise amplification and Gaussian filtering, the 10 Gbps DPSK signal with a power of 6.54 dBm is fed into the MZI and differentially detected by the BPD. Moreover, a voltage of 3.4 V is applied to phase tuning port of MZI for optimum self-differential demodulation which can maximize the output signal amplitude. The eye diagram of final generated 10 Gb/s NRZ electrical signal is very clear with a Q factor of 7.8 measured by the DSO, as shown in Fig. 9(f). Similarly, eye diagrams for OOK and DPSK signals with other data rates are also measured and shown in Fig. 9. Intuitively, these signals have high Q factor.
Figures 10(a)-10(c) show the measured performance parameters of demodulated OOK and DPSK signals at different data rates. The parameters correspond to extinction ratio (ER), signal-to-noise ratio (SNR) and root-mean-square (RMS) timing jitter, respectively. In order to make the input conditions of the OOK and DPSK signals at different data rates the same, the power entering the C-band receiver is adjusted to −33 dBm by adjusting the VOA. The ERs and SNRs of demodulated OOK and DPSK signals with data rates from 1 Gbps to 10 Gbps, are distributed in the range of 7.3 to 10.1 dB and 14.8 to 19.9 dB. Moreover, when the data rate increases from 1 Gbps to 10 Gbps, about 2.1 (1.9) dB ER loss and 3.9 (3.6) dB SNR loss will arise for OOK (DPSK) signal. The RMS time jitter fluctuates between 6 and 12 ps and is not proportional to the data rate as shown in Fig. 10(c).
In the end, the BERs are measured to quantitatively estimate the quality of the regenerated signal. Figure 11(a) shows the BERs as functions of the received power, for the 10 Gbps OOK and DPSK signals, and two BTB signals for comparison. As shown in Fig. 11(a), the required powers for the 10 Gbps OOK and DPSK are −37.12 dBm and −39.24 dBm at BER of 1 × 10−6, while the required power for the two BTB signals are −40.13 dBm and −42.36 dBm, respectively. Therefore, because of noise introduced in the conversion process, such as amplifier noise and nonlinear noise, the power penalties are 3.01 dB for the OOK signal and 3.12 dB for the DPSK signal. Figure 11(b) shows the BERs of DPSK signal at different data rates as functions of receiver power, and Fig. 11(c) corresponds to the BERs of OOK signal. For DPSK signal with data rates from 1 to 10 Gbps, the required powers increase from −42.86 to –39.24 dbm at BER of 1 × 10−6. Similarly, for OOK signals, the required powers increase from −41.3 to −37.12 dBm.
4. Summary and conclusions
In conclusion, we have built a C-band–MIR–C-band FSO link with a speed of 10 Gbps, and demonstrate the successful down conversion and up conversion between the 1550 nm and the 3594 nm based on DFG by using robust MIR transmitter and receiver modules. To the best of our knowledge, for the first time to achieve MIR transmission at this high data rate. The tunable MIR transmitter generates 5.34 dBm power at 3594 nm and the MIR receiver regenerates −24.5 dBm power at 1550 nm. The demonstrated DFG-based MIR system has shown high stability performance of transmission for all tested signals. Both for OOK and DPSK modulation formats, the power penalty compared with BTB signals are lower than 3.5 dB measured at BER of 1E-6. The experimental results have shown that the system is transparent to variable data rate, modulation format and wavelength. Furthermore, the well-integrated and portable MIR transmitter and receiver modules can be applied to field experiments.
National Key R&D Program of China (2017YFC0803900); National Natural Science Foundation of China (NSFC) (No.9163801).
1. M. Grabner and V. Kvicera, “Multiple Scattering in Rain and Fog on Free-Space Optical Links,” J. Lightwave Technol. 32(3), 513–520 (2014). [CrossRef]
2. V. W. S. Chan, “Free-Space Optical Communications,” J. Lightwave Technol. 24(12), 4750–4762 (2006). [CrossRef]
3. K. Li, J. Ma, L. Tan, S. Yu, and C. Zhai, “Performance analysis of fiber-based free-space optical communications with coherent detection spatial diversity,” Appl. Opt. 55(17), 4649–4656 (2016). [CrossRef] [PubMed]
4. B. Janjua, H. M. Oubei, J. R. Durán 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]
5. 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]
6. B. Janjua, T. K. Ng, C. Zhao, H. M. Oubei, C. Shen, A. Prabaswara, M. S. Alias, A. A. Alhamoud, A. A. Alatawi, A. M. Albadri, A. Y. Alyamani, M. M. El-Desouki, and B. S. Ooi, “Ultrabroad linewidth orange-emitting nanowires LED for high CRI laser-based white lighting and gigahertz communications,” Opt. Express 24(17), 19228–19236 (2016). [CrossRef] [PubMed]
7. D. V. Murphy, M. I. Lincoln, D. M. Boroson, J. J. Scozzafava, and B. S. Robinson, “The Lunar Laser Communications Demonstration (LLCD),” 2009 Third IEEE International Conference on Space Mission Challenges for Information Technology, Pasadena, CA, 23–28 (2009).
8. M. Gregory, F. F. Heine, H. Kampfner, R. Lange, M. Lutzer, and R. Meyer, “Commercial optical inter-satellite communication at high data rates,” Opt. Eng. 51(3), 031202 (2012). [CrossRef]
9. S. Seel, H. Kampfner, F. Heine, D. Dallmann, G. Muhlnikel, M. Gregory, M. Reinhardt, K. Saucke, J. Muckherjee, U. Sterr, B. Wandernoth, R. Meyer, and R. Czichy, “Space to Ground bidirectional optical communication link at 5.6 Gbps and EDRS connectivity outlook,” 2011 Aerospace Conference, MT, 1–7 (2011). [CrossRef]
10. T. Plank, E. Leitgeb, P. Pezzei, and Z. Ghassemlooy, “Wavelength-selection for high data rate Free Space Optics (FSO) in next generation wireless communications,” 2012 17th European Conference on Networks and Optical Communications, Spain, 1–5 (2012). [CrossRef]
11. A. Arnulf, J. Bricard, E. Curé, and C. Véret, “Transmission by haze and fog in the spectral region 0.35 to 10 microns,” J. Opt. Soc. Am. 47(6), 491–498 (1957). [CrossRef]
12. A. Soibel, M. W. Wright, W. H. Farr, S. A. Keo, C. J. Hill, R. Q. Yang, and H. C. Liu, “Midinfrared Interband Cascade Laser for Free Space Optical Communication,” IEEE Photonics Technol. Lett. 22(2), 121–123 (2010). [CrossRef]
13. P. Corrigan, R. Martini, E. A. Whittaker, and C. Bethea, “Quantum cascade lasers and the Kruse model in free space optical communication,” Opt. Express 17(6), 4355–4359 (2009). [CrossRef] [PubMed]
15. A. Soibel, M. Wright, W. Farr, S. Keo, C. Hill, R. Q. Yang, and H. C. Liu, “Free space optical communication utilizing mid-infrared interband cascade laser,” Proc. SPIE 7587. Free-Space Laser Communication Technologies XXII, 75870S (2010). [CrossRef]
16. H. Li, W. J. Wan, Z. Y. Tan, Z. L. Fu, H. X. Wang, T. Zhou, Z. P. Li, C. Wang, X. G. Guo, and J. C. Cao, “6.2-GHz modulated terahertz light detection using fast terahertz quantum well photodetectors,” Sci. Rep. 7(1), 3452 (2017). [CrossRef] [PubMed]
17. R. Martini, R. Paiella, C. Gmachl, F. Capasso, E. A. Whittaker, H. C. Liu, H. Y. Hwang, D. L. Sivco, J. N. Baillargeon, and A. Y. Cho, “High-speed digital data transmission using mid-infrared quantum cascade lasers,” Electron. Lett. 37(21), 1290 (2001). [CrossRef]
18. R. Martini, C. Gmachl, J. Falciglia, F. G. Curti, C. G. Bethea, F. Capasso, E. Whittaker, R. Paiella, A. Tredicucci, A. Hutchinson, D. L. Sivco, and A. Y. Cho, “High-speed modulation and free-space optical audio/video transmission using quantum cascade lasers,” Electron. Lett. 37(3), 191 (2001). [CrossRef]
19. X. Pang, O. Ozolins, R. Schatz, J. Storck, A. Udalcovs, J. R. Navarro, A. Kakkar, G. Maisons, M. Carras, G. Jacobsen, S. Popov, and S. Lourdudoss, “Gigabit free-space multi-level signal transmission with a mid-infrared quantum cascade laser operating at room temperature,” Opt. Lett. 42(18), 3646–3649 (2017). [CrossRef] [PubMed]
20. C. Erny, K. Moutzouris, J. Biegert, D. Kühlke, F. Adler, A. Leitenstorfer, and U. Keller, “Mid-infrared difference-frequency generation of ultrashort pulses tunable between 3.2 and 4.8 microm from a compact fiber source,” Opt. Lett. 32(9), 1138–1140 (2007). [CrossRef] [PubMed]
21. M. Asobe, O. Tadanaga, H. Miyazawa, Y. Nishida, and H. Suzuki, “Multiple quasi-phase-matched LiNbO3 wavelength converter with a continuously phase-modulated domain structure,” Opt. Lett. 28(7), 558–560 (2003). [CrossRef] [PubMed]
22. L. Goldberg, W. K. Burns, and R. W. Mcelhanon, “Wide acceptance bandwidth difference frequency generation in quasi-phase-matched LiNbO3,” Appl. Phys. Lett. 67(20), 2910–2912 (1995). [CrossRef]
23. Q. Hao, G. Zhu, S. Yang, K. Yang, T. Duan, X. Xie, K. Huang, and H. Zeng, “Mid-infrared transmitter and receiver modules for free-space optical communication,” Appl. Opt. 56(8), 2260–2264 (2017). [CrossRef] [PubMed]
24. S. Zlatanovic, J. S. Park, S. Moro, J. M. C. Boggio, I. B. Divliansky, N. Alic, S. Mookherjea, and S. Radic, “Mid-infrared wavelength conversion in silicon waveguides using ultracompact telecom-band-derived pump source,” Nat. Photonics 4(8), 561–564 (2010). [CrossRef]
25. K. D. Büchter, H. Herrmann, C. Langrock, M. M. Fejer, and W. Sohler, “All-optical Ti:PPLN wavelength conversion modules for free-space optical transmission links in the mid-infrared,” Opt. Lett. 34(4), 470–472 (2009). [CrossRef] [PubMed]
26. E. Ip, D. Buchter, C. Langrock, J. M. Kahn, H. Herrmann, W. Sohler, and M. M. Fejer, “QPSK transmission over free-space link at 3.8 μm using coherent detection with wavelength conversion,” 34th European Conference on Optical Communication, Brussels, 1–2, (2008).
27. A. V. Okishev and J. D. Zuegel, “Intracavity-pumped Raman laser action in a mid IR, continuous-wave (cw) MgO:PPLN optical parametric oscillator,” Opt. Express 14(25), 12169–12173 (2006). [CrossRef] [PubMed]
29. R. L. Sutherland, D. G. Mclean, and S. Kirkpatrick, Handbook of Nonlinear Optics (Marcel Dekker, 2003), Chap. 2.
30. O. Gayer, Z. Sacks, E. Galun, and A. Arie, “Temperature and wavelength dependent refractive index equations for MgO-doped congruent and stoichiometric LiNbO3,” Appl. Phys. B 91(2), 343–348 (2008). [CrossRef]
32. J. J. Zondy, “The effects of focusing in type-I and type-II difference- frequency generations,” Opt. Commun. 149(1-3), 181–206 (1998). [CrossRef]
33. J.-J. Zondy, D. Touahri, and O. Acef, “Absolute value of the d36 nonlinear coefficient of AgGaS2: prospect for a low-threshold doubly resonant oscillator-based 3:1 frequency divider,” J. Opt. Soc. Am. B 14(10), 2481–2497 (1997). [CrossRef]
34. D. Richter and P. Weibring, “Ultra-high precision mid-IR spectrometer I: Design and analysis of an optical fiber pumped difference frequency generation source,” Appl. Phys. B 82(3), 479–486 (2006). [CrossRef]