Open Access
Add to BibSonomy
Abstract
Plastic optical fiber communication (POFC) systems are particularly sensitive to signal performance and power budget. In this paper, we propose what we belive to be a novel scheme to jointly enhance the bit-error-ratio (BER) performance and coupling efficiency for multi-level pulse amplitude modulation (PAM-M) based POFC systems. The computational temporal ghost imaging (CTGI) algorithm is developed for PAM4 modulation for the first time to resist the system distortion. The simulation results reveal that enhanced BER performance and clear eye diagrams are acquired by using CTGI algorithm with an optimized modulation basis. Experimental results also investigate and show, with CTGI algorithm, the BER performance for 180 Mb/s PAM4 signals is enhanced from 2.2 × 10−2 to 8.4 × 10−4 over 10 m POF by using a 40 MHz photodetector. The POF link is equipped with micro-lenses at its end faces by using a ball-burning technique, which helps to increase the coupling efficiency from 28.64% to 70.61%. Both simulation and experimental results show that the proposed scheme is feasible to achieve a cost-effective and high-speed POFC system with short reach.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Like fiber-to-the-home, fiber-in-the-home (FITH) has also played widespread attention in recent years, which is mainly driven by two major factors: increasing demands for bandwidth and sensitivity to network cost [1]. Such broadband access demand comes from the expected high-quality services of indoor network, such as high-definition (HD) IPTV, multi-room/multi-vision configuration, video conference, CCTV surveillance and so on. Besides, the cost of in-home network is always controlled at a minimum level, due to the large number of network interfaces and fewer users share equipment and installation. Large-core plastic optical fiber (POF) is regarded as one of the most cost competitive candidate in the in-home networks [2,3]. Intensity modulation direct detection (IM/DD) is a practical solution for short-reach optical transmission systems including FITH networks, due to its low complexity, low cost, and low power consumption [4,5]. Advanced modulation formats, such as pulse amplitude modulation (PAM), carrier-less amplitude and phase modulation [6], and orthogonal frequency division multiplexing [7] are feasible to improve spectral efficiency of IM/DD systems. Among them, 4-level PAM (PAM4) is the most powerful modulation format and has been experimentally demonstrated at beyond 200 Gb/s data rate per wavelength [5,8–10].
On one hand, to achieve high-speed capacity in an IM/DD system, it is easy to cause inter-symbol interference (ISI) when transmitting signals at frequencies beyond the limitation of system bandwidth [11]. Moreover, PAM4 signal has low noise tolerance than on-off-keying (OOK) signal. To reduce bit-error-ratio (BER) in the PAM4 system, pre-compensation at the transmitter and post-equalization at the receiver are usually required by using digital signal processing (DSP) [12–14]. Recently, a novel DSP algorithm called computational temporal ghost imaging (CTGI) has been extended from the imaging field to the optical communication field [15–17], which has better robustness to noise [17,18]. Ghost imaging (GI) is an indirect imaging method [19–21], which is used to reconstruct the image of an object by correlating the intensity of reference beam and test beam. Computational ghost imaging (CGI) calculates the distribution of light intensity through prior modulation and eliminate the reference light path [22,23], which greatly simplifies the imaging system and improves the image quality. CTGI technique is the application of CGI in time domain, originally proposed by F. Devax in 2016 [24]. Later, the application of CTGI in the acoustic field was reported [25]. Recently, researchers applied CTGI to the underwater optical communication system with OOK modulation and ensured that low-bandwidth photodetectors can detect high-frequency transmission signals [15]. In [16], a new VLC scheme based on CTGI was successfully implemented, which enables signal transmission at an ultra-high frequency much higher than the bandwidth of the emitter. In [17], an optical encryption scheme for visible light communication is proposed based on TGI with a micro-light-emitting diode. These works focus on transmitting high frequency signal and high security VLC over free space channel. However, the application of CTGI in a PAM4-based optical communication system to reduce POF channel ISI have not been demonstrated yet.
On the other hand, the efficient coupling of POF to light source has become a thorny problem to the power budget in the IM/DD system, due to the lack of cost-effective and easy-installing coupling devices compatible with large-core POF. Therefore, it is crucial to develop a simple and efficient POF coupling scheme. Compared with the external-lens coupling method commonly used in traditional optical coupling systems, micro-lens is simple, compact, assembly-free, easy to package and can be calibrated in one step coaxial. J. Chandrappan’s team used hot blowing and polymer dip fabrication techniques to make the lensed tips, whose coupling efficiency is improved by 27% [26]. Y. Tseng et al. verified an aspherical plastic lens, which was bound with a flattened plastic fiber end by laser transmission welding (LTW) to form an aspherical fiber end face [27]. The coupling efficiency of this method is up to 72%, which is better than that of using spherical lens [28]. J. Huang et al. found that the CO2 laser cutting can be employed to create a micro-lens in POF end to optimize the light gathering or light distribution in textile fields [29]. Considering the practical problems such as cost, easy control and mass production, methods of polishing [30] and chemical etching [31] to produce a lensed end in glass optical fiber are difficult to be extended to the POF field.
In this paper, we proposed an innovative scheme to enhance the transmission performance for PAM4-based plastic optical fiber communication (POFC) system by combining CTGI algorithm and micro-lens coupling technique for the first time. The proposed micro-lens structure can be automatically produced by a CO2 laser fusion machine with pre-set parameters to achieve the maximum coupling efficiency of 70.61%. As for PAM4 and PAM8 signals, it is found that the proposed algorithm is highly effective for noise suppression. As a result, the BER performance is enhanced by about three order of magnitude when compared without compensation. We also implemented the CTGI algorithm for a 180 Mb/s optical PAM4 experimental system over 10 m POF by using a 40 MHz photodetector. The experimental results for PAM4 signal show that, the BER performance is enhanced from 2.2 × 10−2 to 8.4 × 10−4 at the received optical power (ROP) of −12 dBm. Our proposed scheme provides the potential to achieve high-performance and low-cost PAM4-based optical transmission systems.
2. Principle
To build a high-speed POFC system, it usually suffers from penalties from two aspects, including coupling loss between laser source and POF, and ISI caused by POF link as shown in Fig. 1. These two kinds of penalties can be reduced individually. Here, we propose to use micro-lens technique to reduce the coupling loss and CTGI algorithm to compensate the channel ISI. Thereby, the total POFC system performance is significantly improved.
Fig. 1. Typical POFC system in the presence of coupling loss and channel ISI. LD: laser diode, ISI: inter-symbol-interference, POF: plastic optical fiber, APD: Avalanche photodiode.
2.1 Theory of computational temporal ghost imaging (CTGI)
GI mainly reconstruct the target image via the intensity correlation between reference beam and test beam. The reference beam is directly recorded by a spatial resolution detector and the test beam is passed through the object and detected by a non-spatially distributed bucket detector. CGI calculates the distribution of light intensity through prior modulation, which greatly simplifies the imaging system and avoids introducing additional noise by removing the reference optical path. CTGI is the application of CGI in time domain, which is capable of reconstructing a temporal object by calculating the second-order intensity correlation between each group of random illumination patterns generated with a computer and the total light value which is collected by a bucket detector. When extend to the field of optical communication, the process of signal reconstruction can be expressed by matrix operation [15,24]. The original signal (OS) corresponding to the temporal object can be described as a $1 \times L$ vector $\textrm{P} = [{{P_1},{P_2}, \cdots {P_L}} ]$. We chose a $N \times N$ Hadamard matrix M as the modulation basis, corresponding to the random illumination mode, which can help achieve perfect reconstruction [32]. The time-resolved reference intensity in the ${i_{th}}$ measurement can be denoted by a $1 \times N$ vector ${M_i}$ $({\textrm{i} = 1,2, \cdots ,\textrm{N}} )$. Thus, the result of bucket measurements in the ${i_{th}}$ measurement can be expressed as
where T denotes the transpose of the matrix. Finally, the reconstructed signal (RS) can be computed by the second-order correlation function, which is defined asSince the given M is an orthogonal matrix, ${M^T}M$ is a matrix with the same diagonal elements. Thus, the RS can be realized error-free reproduction under an ideal condition.
CTGI algorithm can reduce crosstalk between sampling sequences due to the orthogonality of the modulation basis. Moreover, the reconstruction efficiency is enhanced as increasing the matrix dimension N. At the same time, the algorithm complexity is also involved. When discussing the algorithm computational complexity, the main concern is its approximate value (asymptotic trend), which is often expressed by “O” notation. In the process of implementing CTGI algorithm, the program mainly contains two kinds of complexity, namely linear complexity and constant complexity. We denote the total execution time of an algorithm as $\textrm{T}(n )$. There have $\textrm{T}(n )= \textrm{O}({f(n )} )$, the “O” notation in the formula indicates that the total execution time $\textrm{ T}(n )$ is proportional to the total lines of code $f(n )$, which indicates the asymptotic time complexity of the algorithm, does not mean the actual execution time. Where n is the method parameter, representing the data scale of the algorithm. For simplicity, we ignore the addition constants and subordinates except the highest-order term, and remove the coefficients for the highest-order term. We can describe the time complexity of the proposed algorithm with $\textrm{O}(N )$, where N is the dimension of modulation basis. Without CTGI compensation, the time complexity can be represented with $O(1 )$. Therefore, the selection of N value is a tradeoff between complexity and performance.
The simulation of CTGI-based transmission system is built up and the DSP flow is shown in Fig. 2. At the transmitter side, a data frame with 100,000 pseudo-random binary sequences (PRBS) is generated offline using Matlab. Then the binary data is mapped to PAM-M signal, as OS. Note that the OS is preceded by a preamble of 1000 PRBS that is used for synchronization at the receiver. Subsequently, the generated PAM-M signal is modulated using CTGI algorithm and then up-sampled by 3 times. At the receiver side, the signal is re-sampled after synchronization. Next, we obtained the RS with CTGI reconstruction. To obtain the received binary data for BER calculation, the RS requires PAM-M de-mapping.
To demonstrate the validity of the proposed algorithm in terms of noise suppression, we applied CTGI technique for PAM-M (M = 2, 4, 8) signals, and list two Hadamard matrix dimension N for analysis purpose, respectively. In Fig. 3, the BER performance with and without CTGI technique is compared with the theoretical results, and the corresponding root mean square errors (RMSEs) are also calculated. It is obvious that the BER performance enhanced and the RMSEs decreased as increasing N value. Moreover, the superiority of the proposed technique is more notable at high transmitted signal-to-noise ratio (SNR) region. Figure 3(a) shows the numerical BER results for OOK signals with blue curves. When N increases to 4, the best BER performance achieves, which is close to the theoretical results. At the same time, the BER value with CTGI algorithm is 6.1 × 10−5 at the SNR value of 9 dB. While without CTGI compensation, the BER performance deteriorates to 2.3 × 10−2. When CTGI algorithm is applied for PAM4 signals, considering the tradeoff between computational complexity and BER performance, the values of N are set to 4 and 8. Based on the simulation results as shown in the blue curve of Fig. 3(b) that with CTGI algorithm compensation with N = 8, the achieved BER is 2.5 × 10−5 at the SNR value of 14 dB. While the BER value decreases to 1.3 × 10−2 without CTGI compensation, which is worse than 7% FEC threshold of 3.8 × 10−3. Here, it can be inferred that the needed SNR value to get 7% FEC threshold is about 17 dB for PAM4 signal without CTGI algorithm as shown in Fig. 3(b) of blue curve. Similarly, for PAM-8 signals, when N and SNR are set to 16 and 18, the measured BER results is 1.5 × 10−5 as shown in Fig. 3(c) of blue curve. While the BER value is 4.2 × 10−2 without CTGI algorithm. Besides, when set N to 12, 2 dB SNR penalty is achieved compared with the theory curve at the BER level of 6.3 × 10−6. As shown in Fig. 3 of orange curve, the RMSE of demodulated PAM-M signal with CTGI algorithm is much smaller than that of without CTGI algorithm, and further decreases as increasing N, which indicates smaller penalty by using CTGI compensation. CTGI algorithm contributes to filter out the irrelevant multiplicative noise by calculating the second-order intensity correlation between the modulation basis and the received signals [15]. In the reconstruction process of correlation calculation, only the overall intensity of the detection is needed. The total detection value remains relatively stable even if the signal is distorted by noise during transmission, which does not affect the second-order correlation. Therefore, CTGI algorithm is robust to ISI. Moreover, the correlation is strengthened as increasing the matrix dimension N.
Fig. 3. The numerical BER and root mean square error (RMSE) results for PAM-M signals with and without CTGI algorithm as a function of transmitted signal-to-noise ratio (SNR).
The recovered eye diagrams of PAM-2/4/8 signal with and without CTGI technique under the listed conditions are shown in Fig. 4. Without CTGI algorithm, the PAM-M signal is distorted seriously, and the corresponding eye diagrams in the first row of Fig. 4 are almost closed. While using the proposed algorithm as we can see in the second row of Fig. 4, the eye diagrams are clearly opened, and a relatively uniform level spacing can be observed, which indicates that the ISI is dramatically removed at this time. The proposed CTGI algorithm helps higher-order PAM format signals reduce SNR loss especially with an optimized modulation basis.
Fig. 4. Recovered eye diagrams of normalized amplitude for PAM-M signals with and without CTGI algorithm. (a) N = 4, SNR = 9; (b) N = 8, SNR = 14 and (c) N = 16, SNR = 18.
2.2 Structure design and coupling efficiency analysis
Figure 5 shows the details of the proposed micro-lens structure, which is the combination of aspherical surface and truncated cone. To demonstrate the coupling ability of the proposed micro-lens, we first theoretically analyze the size of the focused spot after the paraxial Gaussian beam passing through the lens. The minimum spot passing through the lens is determined by the refractive index n of the lens, the thickness of the lens ${d_1} + {d_2}$, equivalent focal length f, the wavelength of the incident beam λ and the beam waist radius ${\omega _1}$ at the input plane. The input beam radius onto the POF proximal end is defined as
In the model field, the measured lens far-field angles along the horizontal and vertical directions ${\theta _x}$ and ${\theta _y}$, were used to calculate the laser mode-field radii at the front facet, ${\omega _x}$ and ${\omega _y}$. The relationship between the far-field angle and the mode-field radius at the waist can be derived as
The similar expression can be obtained for ${\omega _y}$. Based on Fresnel coupling theory, the propagation of the Fresnel diffracted beam in free space is also established at the proximal end of POF. The formula of the modal field propagation at the x-direction at the Z distance is
The propagation formula of the modal field at the y-direction $\mathrm{\psi }({\textrm{y},z} )$ is similar to $\mathrm{\psi }({x,z} )$. Thus, the total modal field of the special mode can be expressed as
The modal field of the fundamental mode propagating in the fiber is ${\psi _\textrm{f}}(x,y$). The coupling efficiency $\eta $ is derived from the overlapping integral of the new mode field and the fundamental mode field. Consequently, we have
The change of the incident beam waist ${\omega _2}$ in Eq. (4) or the $\theta $ in the Eq. (5) results in the change of the modal field of the propagating Gaussian beam in Eq. (6). Thus, the coupling efficiency of the interaction between the new mode field and the fundamental mode field changes in Eq. (8), leading to an increase in coupling efficiency.
3. Experimental results and discussions
3.1 Micro-lens fabrication and coupling efficiency measurement
We use a CO2 laser fusion machine (Fujikura, LZM-100) to manufacture the proposed micro-lens. It is equipped with a high-stable CO2 laser as the heating source, with the highest output power of 30 W (900 bits) and the standard power of 413 bits. The output spot energy shows a Gaussian distribution, and the half width of the spot energy is about 300 microns after a lens is installed. After passing through the beam splitter mirror, the emitted light of CO2 laser is divided into two lasers with equal energy, which illuminate the optical fiber. The angle between the two lasers is about 170 degrees. The LZM-100 has two translation motors (Z motors), two rotating motors (θ motors) and two cameras for monitor the manufacturing process in real time. Sensor parameter setting and manufacturing process are controlled by a user-defined specific program. The operation process mainly includes the preparation stage and the formal machining stage.
Preparation stage: Polish the end face of the selected POF (fiber diameter 1000 µm, core 200 µ$\textrm{m}$) with fine sandpaper until it is flat. After washing the end face, we wipe the end face with alcohol cotton and observe its flatness with a microscope, which ensures for no impurities. We then reset the fusion splicer, and then clamp the optical fiber with a 1 mm clip and put it into the fiber fusion splicer at a suitable position for next stage.
Machining stage: Set working parameters on the software that controls the fusion splicer. For optimal working parameters, the diameter of the optical fiber and the burning ball are set to 1000 µ$\textrm{m}$ and 1200 µ$\textrm{m}$, the power of the CO2 laser is set to 60 bits, the rotation speed and the feeding speed are set to 230 reg/s and 0.01 mm/s. We then write into the initialization parameters, and start the process of burning the ball for about 5 minutes.
Subsequently, we measured the coupling efficiency of the finished products. A 650 nm red laser diode (LD) is selected as the light source, which is powered by DC source. The LD is packaged with an optical fiber pigtail and directly connected with a fiber collimator. The collimated beam is further focused by a focusing lens and then coupled into the micro-lens POF. A optical power meter is used to measure the input optical power ${P_0}$ (head side) and output optical power ${P_t}$ (tail side), respectively. Here, the transmission loss of light in the selected short POF is negligible. Finally, the coupling efficiency is directly computed by $\mathrm{\eta } = {P_t}/{P_o}$.
Several groups of micro-lenses are acquired by controlling the working parameters of the machine for many times. Here, we selected 9 groups of representative micro-lenses, whose micrographs and corresponding coupling efficiencies are shown in Fig. 6. Detailed working parameters and corresponding measured mean values are shown in Table 1. Among them, the rotation speed is set as the maximum speed (230 reg/s) to ensure a completely symmetrical structure. According to the results in Table 1, the parameters of the third group (No. 7, 8, 9) are the best parameters, which are determined by comparing coupling efficiencies and spot quality at the exit end. We found that the coupling efficiency of the structure with optimal parameters can reach 69.38% on average. Moreover, we selected the micro-lens with the best performance among the three groups of parameters, respectively. Then compared them with ordinary POF in terms of coupling efficiency as we can see in Table 2. With the proposed micro-lens structure, the achieved coupling efficiency is enhanced from 28.64% to 70.61%. Compared with LTW method, our method can achieve the same magnitude coupling efficiency, while reduce manufacturing configuration and operational complexity. To further verify the general feasibility of this structure, we manufactured 5 m and 10 m micro-lens POFs with the optimal parameters. By removing the insertion loss, the achieved coupling efficiency is also about 70%.
Fig. 6. Micrographs and corresponding coupling efficiencies for micro-lens POFs with different parameters. The corresponding fabrication parameters are listed in Table 1.
Table 1. Average coupling efficiency under different working parameters.
Table 2. Comparison of coupling efficiency for POFs with and without micro-lens.
3.2 Experimental data transmission using micro-lens and CTGI
To further verify the feasibility of the proposed scheme and investigate the performance of CTGI-based optical PAM4 signal, we carried out an experiment with 10 m POF. The experimental setup is shown in Fig. 7. At transmitter side, the offline generated PAM4 data was modulated by CTGI technique, then loaded into an arbitrary waveform generator (AWG) (Tektronix, AWG7122C), whose output amplitude is set to 1 Vpp. The radio frequency signal from the AWG and the drive current provided by a direct current (DC) source (Keithley, 2231A-30-3) were coupled to drive a 650 nm red LD via a Bias-Tee (Mini-circuits, ZFBT-6GW-FT+). A focusing lens with a focal length of 10 cm was placed at front of the LD to collimate the beam. At the focal point of receiver side, the beam is collected by an avalanche photodiode (APD) with 3-dB bandwidth of 40 MHz (Hamamatsu, C12702-12). A circular variable neutral density filter between the focusing lens and 10 m POF is rotated to change the received optical power (ROP). The total system frequency response after 10 m POF link is shown in Fig. 8(a), which depicts the −3dB bandwidth of 22 MHz. The output electrical signal from the APD was captured by a digital signal oscilloscope (Agilent, DSO-X 92004A), then imported into a computer for offline processing, which mainly includes synchronization, re-sampling, reconstruction and de-mapping.
Fig. 7. The experimental setup for 10 m POF transmission system based on CTGI. AWG: arbitrary waveform generator, LD: laser diode, CVND: circular variable neutral density, POF: plastic optical fiber, APD: avalanche photodiode, DSO: digital signal oscilloscope.
Fig. 8. (a) Frequency response of the 10 m POFC system, and (b) The measured BER and PSNR performance of PAM4 signal for CTGI-based POFC system with and without micro-lens at different data rates, when N is 8.
To validate the feasibility of the micro-lens structure to the POFC system, we firstly carry out an experiment to evaluate the BER and peak signal-to-noise ratio (PSNR) performance of PAM4 signal over 5 m POF link at different data rates as shown in Fig. 8(b). The modulation matrix dimension N is set to 8. The incident optical power to POF end face remains stable and constant. We obtain the outgoing optical power from POF with and without micro-lens is −8 dBm and −12 dBm, respectively, which means 4 dB coupling loss reduction. When the transmission data rate is fixed at 140 Mb/s, the PSNR of PAM4 signal is increased from 12.9 dB to 14.5 dB by using micro-lens coupling, which acquired 1.6 dB enhancement. Meanwhile, the corresponding BER performance is enhanced from 1.5 × 10−4 to 5 × 10−5. This is because the micro-lens improves the coupling efficiency between light source and POF link, leading to an increase in the ROP.
Figure 9(a) describes the measured BER results at various ROPs at 180 Mb/s data rate. With ROP ranging from −18 dBm to −12 dBm, the BER performance of the PAM4 signal without CTGI compensation can be hardly improved due to strong ISI. Moreover, the BER performance is enhanced as increasing the ROP via using the proposed algorithm, especially at high ROP region. When ROP increases to −12 dBm, the BER value of blue curve that without CTGI compensation is 2.3 × 10−2, which is below 7% FEC threshold of 3.8 × 10−3. Meanwhile, the BER value of red curve that with CTGI compensation is enhanced to 8.5 × 10−4. As shown in Fig. 9(b), we select the data sequences with length of 100 from OS and standardized RS for comparison, corresponding to the ROP is −12 dBm in Fig. 9(a) of red curve. It can be seen the weak distortion of RS, which is almost a perfect representation of OS.
Fig. 9. (a) BER performance of 180 Mb/s PAM4 signals as a function of ROP and (b) Part of the data selected from OS and RS when ROP is −12 dBm and N is 8.
When ROP is fixed at – 18 dBm, the BER results under different transmission rates over 10 m POF are shown in Fig. 10. The PAM4 signal without CTGI compensation at data rate of 130 Mb/s achieves the BER level of 4.5 × 10−3. While with CTGI compensation the achieved data rate is significantly enhanced. Attentively, the transmitted baud rate of PAM4 signal in the experiment is 90 MBaud/s, which is higher than the 3-dB bandwidth of the used APD module of 40 MHz. Thereby, the PAM4 IM/DD communication system without CTGI compensation suffers from severe ISI due to the limited system bandwidth. The impaired signal is hardly improved as increasing the ROP. Fortunately, CTGI technique can effectively alleviate the impact of ISI and has strong robustness to noise especially at high ROP region. It means that compared to direct signal transmission, CTGI technique can obtained the same BER performance at lower ROP. In other words, CTGI technique helps to extend the communication distance or save energy consumption.
Fig. 10. The measured BER results for PAM4 signal over 10 m POF link with and without CTGI algorithm as a function of data rate, when ROP is −18 dBm.
4. Conclusion
In this paper, we have proposed a novel scheme of noise mitigation and fiber coupling optimization for PAM4-based POFC system by using CTGI technique and micro-lens POF. The proposed method effectively addresses the system issues of PAM4-based POFC, which are spectral efficiency and fiber coupling efficiency. We experimentally achieved the maximum coupling efficiency of 70.61% by using ball-burning technique, manufactured by a CO2 laser fusion machine with characteristics of automation and high repeatability. Compared with ordinary POF, the coupling loss is reduced from 5.4 dB to 1.5 dB, which shows 3.9 dB enhancement. To achieve high spectral efficiency, we use advanced multi-level PAM modulation. The simulation results show that the BER performance of PAM-2/4/8 signal can be significantly improved by using the proposed CTGI algorithm. Compared with theory curve, when set N and SNR to 8 and 14, just 1 dB SNR penalty for PAM4 signal at the BER level of 2.5 × 10−5 is successfully achieved. CTGI also shows its superior anti-noise performance for PAM8 signal. The proposed CTGI algorithm is also experimentally implemented for a 180 Mb/s PAM4-based POFC system. The experimental results show that, the BER performance is enhanced from 2.2 × 10−2 to 8.4 × 10−4 at the received optical power (ROP) of −12 dBm, which reveals the feasibility of the proposed scheme.
Funding
National Natural Science Foundation of China (62205166, 11874012, 12174002); Major Key Project of Peng Cheng Laboratory; Excellent Scientific Research and Innovation Team of Anhui Province (2022AH010003); Key Research and Development Plan of Anhui Province (202104a05020059); Innovation Project for the Returned Overseas Scholars of Anhui Province (2021LCX011); The University Synergy Innovation Program of Anhui Province (GXXT-2020-052); Anhui Project (Z010118167).
Disclosures
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
Data underlying the results presented in this Letter are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. B. Liu, M. Peng, W. Wang, and B. Liu, “Fiber in the Home: Broadband Networks and Optic Distributed Sensing,” presented at the 2008 International Seminar on Future BioMedical Information Engineering, 2008. [CrossRef]
2. A. M. J. Koonen, H. P. A. van den Boom, E. Ortego Martinez, A. Pizzinat, Ph. Guignard, B. Lannoo, C. M. Okonkwo, and E. Tangdiongga, “Cost optimization of optical in-building networks,” Opt. Express 19(26), 399–B399 (2011). [CrossRef]
3. Y. Shi, E. Tangdiongga, A. M. J. Koonen, A. Bluschke, P. Rietzsch, J. Montalvo, M. de Laat, G. N. van den Hoven, and B. Huiszoon, “Plastic-optical-fiber-based in-home optical networks,” IEEE Commun. Mag. 52(6), 186–193 (2014). [CrossRef]
4. L. Shu, Z. M. Yu, Z. Q. Wan, and K. Xu, “Application Study of Clipping and Digital Resolution Enhancer for Low-Resolution High-Speed IM/DD PAM4 Transmission,” in Conference on Lasers and Electro-Optics Pacific Rim (CLEO-PR), Electr Network, 2020. [CrossRef]
5. X. Pang, O. Ozolins, R. Lin, L. Zhang, A. Udalcovs, L. Xue, R. Schatz, U. Westergren, S. Xiao, W. Hu, G. Jacobsen, S. Popov, and J. Chen, “200 Gbps/Lane IM/DD Technologies for Short Reach Optical Interconnects,” J. Lightwave Technol. 38(2), 492–503 (2020). [CrossRef]
6. T. J. Zuo, A. Tatarczak, M. Iglesias Olmedo, J. Estaran, J. Bevensee Jensen, Q. Zhong, X. Xu, and I. Tafur Monroy, “O-band 400 Gbit/s Client Side Optical Transmission Link,” in Optical Fiber Communications Conference and Exhibition (OFC), San Francisco, CA, 2014. [CrossRef]
7. Z. C. Wang, T. Q. Mao, and Q. Wang, “Optical OFDM for Visible Light Communications,” in 13th International Wireless Communications and Mobile Computing Conference (IWCMC), Valencia, Spain, 2017, pp. 1190–1194, 2017.
8. K. P. Zhong, X. Zhou, T. Gui, L. Tao, Y. Gao, W. Chen, J. Man, L. Zeng, A. Pak Tao Lau, and C. Lu, “Experimental study of PAM-4, CAP-16, and DMT for 100 Gb/s Short Reach Optical Transmission Systems,” Opt. Express 23(2), 1176–1189 (2015). [CrossRef]
9. S. Yamamoto, H. Taniguchi, M. Nakamura, and Y. Kisaka, “O-Band 10-km PAM Transmission Using Nonlinear-Spectrum-Shaping Encoder and Transition-Likelihood-Based Decoder with Symbol- and Likelihood-Domain Feedbacks,” in European Conference on Optical Communication (ECOC), Bordeaux, FRANCE, 2021. [CrossRef]
10. C. Y. Li, X. H. Huang, H. H. Lu, Y. C. Huang, Q. P. Huang, and S. C. Tu, “A WDM PAM4 FSO-UWOC Integrated System With a Channel Capacity of 100 Gb/s,” J. Lightwave Technol. 38(7), 1766–1776 (2020). [CrossRef]
11. M. Xiang, Z. P. Xing, E. El-Fiky, M. Morsy-Osman, Q. B. Zhuge, and D. V. Plant, “Single-Lane 145 Gbit/s IM/DD Transmission With Faster-Than-Nyquist PAM4 Signaling,” IEEE Photonics Technol. Lett. 30(13), 1238–1241 (2018). [CrossRef]
12. Y. M. Liu, Y. H. Zhang, Y. F. Peng, and Z. Z. Zhang, “Equalization of chromatic dispersion using wiener filter for coherent optical receivers,” IEEE Photonics Technol. Lett. 28(10), 1092–1095 (2016). [CrossRef]
13. X. Z. Tang, Y. J. Qiao, Y. W. Chen, Y. M. Lu, and G. K. Chang, “Digital Pre- and Post-Equalization for C-Band 112-Gb/s PAM4 Short-Reach Transport Systems,” J. Lightwave Technol. 38(17), 4683–4690 (2020). [CrossRef]
14. H. H. Zhou, Y. Li, Y. Liu, L. Yue, C. Gao, W. Li, J. Qiu, H. Guo, X. Hong, Y. Zuo, and J. Wu, “Recent Advances in Equalization Technologies for Short-Reach Optical Links Based on PAM4 Modulation: A Review,” Appl. Sci. 9(11), 2342 (2019). [CrossRef]
15. X. W. Chen, M. Jin, H. Chen, Y. Wang, P. Qiu, X. Cui, B. Sun, and P. Tian, “Computations temporal ghost imaging for long-distance underwater wireless optical communication,” Opt. Lett. 46(8), 1938–1941 (2021). [CrossRef]
16. Y. P. Wang, W. Jiang, H. Chen, X. Chen, Y. Gao, P. Tian, and B. Sun, “Ultra-high frequency signal transmission based on temporal ghost imaging,” J. Lightwave Technol. 40(16), 5426–5431 (2022). [CrossRef]
17. Y. P. Wang, H. Chen, W. Jiang, X. Li, X. Chen, X. Meng, P. Tian, and B. Sun, “Optical encryption for visible light communication based on temporal ghost imaging with a micro-LED,” Optics and Lasers in Engineering 134, 106290 (2020). [CrossRef]
18. Y. K. Xu, S. H. Sun, W. T. Liu, G. Z. Tang, J. Y. Liu, and P. X. Chen, “Detecting fast signals beyond bandwidth of detectors based on computational temporal ghost imaging,” Opt. Express 26(1), 99–107 (2018). [CrossRef]
19. T. B. Pittman, Y. H. Shih, D. V. Strekalov, and A. V. Sergienko, “Optical imaging by means of two-photon quantum entanglement,” Phys. Rev. A 52(5), R3429–R3432 (1995). [CrossRef]
20. R. S. Bennink, S. J. Bentley, and R. W. Boyd, ““Two-photon” coincidence imaging with a classical source,” Phys. Rev. Lett. 89(11), 113601 (2002). [CrossRef]
21. A. Valencia, G. Scarcelli, M. D’Angelo, and Y. Shih, “Two-photon imaging with thermal light,” Physical Review Letters, vol. 94, no. 6, Feb. 2005. [CrossRef]
22. B. Sun, M. P. Edgar, R. Bowman, L. E. Vittert, S. Welsh, A. Bowman, and M. J. Padgett, “3D Computational Imaging with Single-Pixel Detectors,” Science 340(6134), 844–847 (2013). [CrossRef]
23. J. H. Shapiro, “Computational ghost imaging,” Phys. Rev. A 78(6), 061802 (2008). [CrossRef]
24. F. Devaux, P. A. Moreau, S. Denis, and E. Lantz, “Computational temporal ghost imaging,” Optica 3(7), 698–701 (2016). [CrossRef]
25. Y. Tian, H. Ge, X. Zhang, X. Xu, M. Lu, Y. Jing, and Y. Chen, “Acoustic ghost imaging in the time domain,” Phys. Rev. Appl. 13(6), 064044 (2020). [CrossRef]
26. J. Chandrappan, Z. Jing, R. Veppathur Mohan, P. Oliver Gomez, T. Aung, X. Yongfei, P. V. Ramana, J. H. Lau, and D. Lee Kwong, “Optical Coupling Methods for Cost-Effective Polymer Optical Fiber Communication,” IEEE Trans. Compon. Packag. Technol. 32(3), 593–599 (2009). [CrossRef]
27. Y. T. Tseng, S. M. Chang, S. H. Huang, and W. H. Cheng, “Lensed plastic optical fiber with an aspherical fiber end formed by joining an aspherical plastic lens and a plastic optical fiber using laser transmission welding,” Precis. Eng. 35(4), 704–711 (2011). [CrossRef]
28. Y. T. Tseng, J. B. Huang, T. Y. Hung, H. Y. Jhou, and C. L. Kuo, “Lensed plastic optical fiber employing hyperbolic end filled with high-index resin using electrostatic force,” Precis. Eng. 38(1), 183–189 (2014). [CrossRef]
29. J. Huang, J. Wiener, D. Kremenakova, J. Militky, and M. Kasparova, “Micro-lensed polymeric optical fiber by CO2 laser cutting,” J. Laser Appl. 30(3), 032014 (2018). [CrossRef]
30. S. I. E. Lin, “A lensed fiber workstation based on the elastic polishing plate method,” Precis. Eng. 29(2), 146–150 (2005). [CrossRef]
31. H. M. Yang, S. Y. Huang, C. W. Lee, T. S. Lay, and W. H. Cheng, “High-coupling tapered hyperbolic fiber microlens and taper asymmetry effect,” J. Lightwave Technol. 22(5), 1395–1401 (2004). [CrossRef]
32. S. Jiang, X. Li, Z. Zhang, W. Jiang, Y. Wang, G. He, Y. Wang, and B. Sun, “Scan efficiency of structured illumination in iterative single pixel imaging,” Opt. Express 27(16), 22499–22507 (2019). [CrossRef]
