6-m/10-Gbps underwater wireless red-light laser transmission system

Abstract. An underwater wireless red-light laser transmission system using 10-Gbps 16-quadrature amplitude modulation–orthogonal frequency-division multiplexing (OFDM) modulation based on a high-speed multimode 680-nm vertical-cavity surface-emitting laser (VCSEL) was proposed and demonstrated. This study is the first attempt to adopt red light for transmitting a 10-Gbps wireless signal 6-m underwater. With the adoption of the appropriate OFDM base bandwidth and modulation parameters after studying the frequency characteristics of the high speed multimode 680-nm VCSEL and contrast experiment of modulation parameters, a good bit error rate performance and clear constellation maps are achieved.


Introduction
In the past, underwater wireless laser transmission systems adopted a 405-nm blue-laser light for underwater transmission, which has minimum attenuation in clear water. [1][2][3][4][5] Compared with blue-laser light, the absorption rate of redlaser light in clear water is higher, resulting in a shorter transmission distance. However, red laser has a significantly wider frequency response at a much lower price and, thus, provides a considerably wider transmission bandwidth. In addition, in terms of the photodiode (PD), the receiving sensitivity of the 620-to 750-nm red-light bandwidth is far higher than that of the 380-to 495-nm blue-light bandwidth. 6 Moreover, considerable research on underwater wireless light communication has been conducted in recent years. A study on underwater transmission characteristics, attenuation coefficient, and channel bandwidth regarding blue, green, and red lights had been proposed, [7][8] from which an interesting finding was obtained-red light has a better transmission performance compared with blue and green lights in high turbidity water. [9][10][11][12][13][14] Therefore, an underwater wireless red-light laser transmission (UWRLLT) system is expected to provide a short-distance underwater highspeed link, similar to Wi-Fi-on-air in terms of function. In the situation of increasing popularity of oceanic exploration and oceanic leisure activities, the sense of distance experienced underwater or in the ocean is more obvious than that on the land. Whether it is for activity, construction, exploration and tracking, or research and development, underwater communication facility is in urgent need for managing site condition, making UWRLLT extremely attractive to the application of underwater communication.
UWRLLT systems are used to transmit data in water using red laser. In this study, a 10-Gbps 16-quadrature amplitude modulation (QAM)-orthogonal frequency-division multiplexing (OFDM) modulation UWRLLT system based on a high-speed multimode 680-nm vertical-cavity surface-emitting laser (VCSEL) is proposed and experimentally demonstrated. The 16-QAM-OFDM is a modulation by which the information bits are transferred into four lowspeed bit-streams then modulated to perpendicular subcarriers after being encoded and inserting the synchronization and lead element. The system was employed to demonstrate a blue-light laser underwater transmission system. [15][16][17] The 16-QAM-OFDM has a very high-frequency effect because it can transmit nearly four times the amount of data at a low digital rate. To the best of the authors' knowledge, this study is the first attempt to adopt a 680-nm VCSEL transmitter in a 10-Gbps 16-QAM-OFDM UWRLLT system. After studying the frequency response of the system and 16-QAM-OFDM modulation parameters, a performance with a bit rate of 10 Gbps at a bit error rate (BER) of 2.9 × 10 −4 is obtained from a 6-m underwater transmission using the UWRLLT system. BER is a good guarantee of the quality of communication due to its significantly lower value than the forward error correction (FEC) limit of 3.8 × 10 −3 . In contrast to a high-speed underwater blue-laser wireless system, 15 the UWRLLT system has a simpler configuration, lower cost, and lower BER.
2 Experimental Setup Figure 1 shows the configuration of the proposed 10-Gbps 16-QAM-OFDM UWRLLT system that adopts a 680-nm red-light VCSEL transmitter. The VCSEL, with wavelength range/color of 679 to 681 nm∕red, is directly modulated by a 16-QAM-OFDM data stream. An 16-QAM-OFDM data signal generated offline by the MATLAB program is fed into an arbitrary waveform generator (AWG, Tektronix 7102), and pseudorandom bit sequence data streams with a length of 2 15 − 1 are converted into parallel binary data by a serialto-parallel (S/P) module. After being encoded, the signal is modulated onto orthogonal subcarriers. The OFDM signals generated by an inverse fast Fourier transform (IFFT) was added with a cyclic prefix (CP) before the 16-QAM-OFDM data are converted into parallel binary data by the S/P module. The signal is fed into the AWG after being converted into analog signals using a digital-to-analog converter. The 10-Gbps 16-QAM-OFDM signal from the AWG is supplied to the 680-nm VCSEL after a bias tee (Bias-T), by which the VCSEL is directly modulated. The light sent out from the VCSEL is fed into the convex lens 1 and transmitted underwater in a water tank and then coupled into convex lens 2 at the receiving site. The function of convex lens 1 is to generate a collimated light beam, whereas convex lens 2 is to couple the collimated light beam into a point. The work range of wavelength of both two convex lens is 350 to 700 nm. The diameter of the collimated light beam is 0.6 mm with a focal length of 5.12 mm. The size of the water tank is 2 m × 1.5 m × 0.8 m, which is full of tap water with decay factor of 0.074 m −1 and the temperature at 23°C. 18 Several mirrors are placed on the side of both water tanks to reflect light three times for extending underwater light transmission distance over 6 m (3 × 2 m). At convex lens 2, which is the receiving side, the transmitted light signal is received by an avalanche photodetector (APD). The signal from an APD is transmitted into a communication signal analyzer (CSA, Tektronix CSA7404B), whereas the signal data output by the CSA is imported into a computer for demodulation and handled offline via MATLAB processing. After the S/P, analog to digital conversion (ADC) and being subtracted from the CP, the OFDM signals are converted using FFT. For evaluating the BER performance and the corresponding star map, the signal is demodulated by a 16-QAM before the BER value is calculated.
In addition, Fig. 1 shows the measurement setup of the frequency response of the 680-nm VCSEL transmitter in the 10-Gbps 16-QAM-OFDM UWRLLT system. An RF sweep signal (10 MHz to 40 GHz) is generated from a network analyzer and fed into the 680-nm VCSEL. The received RF sweep signal from APD detection is fed into the network analyzer. The frequency response of the 680nm VCSEL shows the measurement result under different transmission distances for the back-to-back (BTB) and UWRLLT systems. Figure 2 shows the frequency response of the 680-nm VCSEL for BTB and the frequency response of the UWRLLT system over different underwater distances. If the sudden drop at 3.8 GHz is ignored, then the 3-dB bandwidth for BTB is 4.7 GHz. However, through a 2-m underwater transmission, the 3-dB bandwidth is at 3.6 GHz before   Figure 3(a) shows the spectrum of the data signal from AWG, with 28 dB between the receiving signal and nearby noise. Figure 3(b) shows the spectrum of the data signal through a 2-m underwater transmission with 18 dB between the receiving signal and nearby noise and 12 dB between them at 4 GHz. With a low-pass filter (LPF) as shown in Fig. 1, the BER is 4.9 × 10 −13 through the 2-m underwater transmission, which is slightly below the BER value (7.2 × 10 −13 ) without employing LPF. However, the LPF was not adopted due to cost consideration. Figure 3(c) shows the spectrum of the data signal through the 6-m underwater transmission with 7 dB between the receiving signal and nearby noise. Although the signal-to-noise ratio decreases with the extension of the underwater transmission distance, the receiving signal remains easy to identify

Experimental Results and Discussion
The inverse transformation formula of the FTT can be expressed as follows: E Q -T A R G E T ; t e m p : i n t r a l i n k -; e 0 0 2 ; 3 2 6 ; 7 5 2 XðkÞW kn N ; k ¼ 0; 1; : : : ; N − 1; (2) where N is associated with the spectral resolution that affects BER. Figure 4 shows the measured BER curves of the 10-Gbps 16-QAM-OFDM signal through the 2-m underwater transmission for the 512-and 256-point FFT. The BER of the 512-point FFT is significantly lower than that of the 256-point FFT. Although the spectral resolution can be improved further by a 1024-point FFT, computational complexity significantly increases, which causes further device complexity and cost. Thus, the 512-point FFT is selected. Figure 5(a) shows the optical output power of the 680-nm VCSEL at different biases. A maximum power of 5.76 mW is obtained when the bias current is increased to 8.5 mA. The lowest BER is measured at a bias current of 6.5 mA. The received power (P) after underwater transmission can be expressed as follows: E Q -T A R G E T ; t e m p : i n t r a l i n k -; e 0 0 3 ; 3 2 6 ; 5 2 9 where P o is the power of transmitted light, d is the light transmission distance, and CðλÞ is the attenuation coefficient. When the bias current is 6.5 mA, the measured P at the 2-    Figure 6 shows the measured BER curves of the 10-Gbps 16-QAM-OFDM data signal versus average received optical power over varying underwater distances. Corresponding constellation maps are included among the curves. The measured BER of the 2-, 4-, and 6-m underwater transmission is 7.2 × 10 −13 , 3.2 × 10 −6 , and 2.9 × 10 −4 , respectively. Even for a BER of 2.9 × 10 −4 , it is significantly lower than the FEC limit of 3.8 × 10 −3 . Good BER performance and constellation maps are obtained to demonstrate the feasibility of establishing a 6-m∕10-Gbps UWRLLT system based on a 680-nm VCSEL.
An even lower BER can be derived by employing LNA and equalisers before the signal is transmitted into CSA 4,15 or using a MATLAB process of equalization before demapping. 14 Considering a lower system cost and a significantly low BER of 2.9 × 10 −4 compared with the FEC limit, the LNA and equalizer are not adopted in this study. In contrast to the 10-Gbps blue light UWLT system, 15 the UWRLLT system has a simpler configuration, lower cost, and lower BER (as listed in Table 1). In the 10-Gbps blue light UWLT system, a 405-nm blue laser diode (LD),  A PD is used along with LNA and equalizer.

Comparison
The system in Ref. 15 has one more LD and PD, which is more complex and of higher cost.
With additional LNA and equalizer, the system in Ref. 15 is more complex and of higher cost.

Same
Under the same condition without LNA and equalizer, the configuration of our system performs better than that in Ref. 15 in respect of distance and BER.