To effectively estimate the optical fiber temperature of distributed optical fiber sensing based on Brillouin scattering, a multi-layer feedforward Artificial Neural Network (ANN) is introduced to estimate the temperature.

The article has developed programs in Matlab for fiber optic temperature calculation using the single slope method, the least squares fitting method based on the pseudo-Voigt model，and an ANN. At the same time, the Brillouin spectra with different Signal-to-Noise Ratios (SNR) are simulated. Based on the Brillouin spectrum generated from the above simulations, the study investigates the key parameters of the ANN, namely the number of hidden layers, the number of neurons in the hidden layers, and the training objectives, and their influence on training speed, temperature calculation time, and accuracy.

The maximum temperature error of ANN is only 1.18 and 0.63 ℃ at 22 and 37 dB respectively, and the calculation time of ANN is only about 1/1 000 of that of the least-squares fit method. When the number of hidden layer neurons remains constant，the training time decreases obviously with the number of hidden layer and the computation time increases linearly with the number of hidden layer. However, it has little effect on the accuracy of temperature estimation. Both the training time and computation time increase with the number of neurons in the hidden layer. When there are 21 neurons in the hidden layer, the training time is approximately 67 times that of only one neuron in the hidden layer. However, it also has little effect on the accuracy of temperature estimation. When the training goal (square of the Brillouin shift error) is less than the critical value (about 1 MHz²), the temperature error is almost independent. However, when the training goal exceeds the critical value, the temperature error increases with the training goal.

When ANN is used to estimate optical fiber temperature of distributed optical fiber sensing based on Brillouin scattering, it is recommended to select a single-hidden-layer ANN and the number of hidden layer neurons is set to one. The training goal is set to 1 MHz².

In order to solve the problem of complex and difficult implementation of large capacity and ultra-long haul transmission system in C＋L band, a Remote Gain Unit (RGU) is designed to amplify both C and L bands.

The article examines the theory of optical signal amplification by Erbium-Doped Fiber (EDF) and derives the relationship between the length of EDF and the gain as well as the noise figure of the RGU. Then, through simulation, the gain and noise factor of the RGU under different design schemes are compared. The best EDF length of C and L bands in RGU is selected, and the optimal performance of the C＋L band RGU based on the Remote Optical Pump Amplifier (ROPA) system is designed by optimizing the parameters of the optical path sub-system. The article also reports on the experimental verification of a single-carrier 400 Gbit/s ultra-large capacity long-span unrepeatered transmission system based on a backward RGU, operating in the C＋L bands.

A new record of 36 Tbit/s (90×400 Gbit/s) repeaterless transmission of 342.5 km is achieved in the experiment by using forward high-order Raman amplification technology and high-order pump combined with backward ROPA amplification technology.

The results show that the gain and noise factor of the experimental C＋L band RGU are highly consistent with the results of the simulation data. The design of the 17 m EDF in the C band and 21 m erbium fiber in the L band under this remote gain structure is consistent with the theoretical research. The transmission records of the experimental system show that the C＋L band RGU designed based on theoretical research is feasible in the non-relay large-capacity transmission system and can greatly improve the transmission distance of the C＋L band transmission system. The experimentally significant results obtained in the article provide experimental evidence for the performance, indicators, parameters, and standards of the system, guiding the expansion of signal bands in the application of high-capacity Ultra-Long-Haul (ULH) systems in China.

In recent years, equalizer based on Neural Network (NN) has been widely used in optical fiber nonlinear impairment compensation. However, in practical application, it needs to consume a lot of resources to retrain NN to adapt to optical communication system in new environment. Transfer learning applies some parameters of the NN model trained by the initial system (source domain) to the NN model in the new environment (target domain). Only a small amount of training data is needed to achieve rapid reconstruction of the target domain model. However, this method needs to find the best source domain in all source domains for migration to obtain good performance. When the target domain changes, it is necessary to find the best source domain again, which will consume a lot of training resources. This work suggests a solution based on multi-source domain transfer learning to solve this issue.

This method employs Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) as equalizers. It alternately updates network parameters through two processes: specific source domain training and multi-source domain training. Subsequently, the optical communication system in the new environment is fine-tuned, allowing it to adapt to changes in the transmission system using only a small amount of initial training data. Moreover, good performance can be achieved without the need to search for the optimal source domain.

A 5-channel 50-GBaud Wavelength Division Multiplexing (WDM) Dual-Polarization 16-order Quadrature Amplitude Modulation (DP-16QAM) optical transmission system is simulated to verify the effectiveness of the proposed method. The numerical simulation results show that the multi-source domain transfer learning outperforms the retraining method when using just 35% of the target domain data. Meanwhile, the Q-factor of multi-source domain transfer learning are improved by 0.82 and 0.18 dB, respectively, in compared with retraining and single source domain transfer learning.

Therefore, the multi-source transfer learning scheme is suitable for practical optical communication systems.

In order to eliminate the influence of the physical characteristics of the intrinsic dark current of the Positive-Negative (PN) junction in the Photoelectric Detector (PD), the detection accuracy of the PD can be improved at the same time of enhancing the detection range of the detector power and widening the detector working environment temperature. In this paper, a set of dark current compensation algorithms is designed to calibrate the PDs.

Firstly, we analyze the physical characteristics of the photodiode and the hardware circuit design of the PD to establish a mathematical model between the photogenerated current and the optical power. Then we correct the curve relationship between the optical power and the Analog Digital Convert (ADC) value of the photogenerated voltage through the mathematical model, so as to achieve the power compensation by removing the influence of the dark current power at the specific temperature point. Next, by studying the temperature characteristics of dark current, we calculate the power required to compensate for dark current under a limited number of temperature conditions. By fitting these few temperature points, we extend the compensation to a broader temperature range, thereby achieving dark current temperature compensation. Lastly, a temperature-controlled testing platform is established, utilizing an Optical Switch (OS) to regulate the incident optical power on the photodiode. The real-time incident optical power on the PD is recorded through an optical power meter. This setup allows for testing the accuracy of the PD under various temperature conditions and incident optical power levels.

The experiments show that the working temperature can be widened from －5～55 ℃ to －40～80 ℃ and the minimum detection power can be reduced from －40 to －68 dBm by using the algorithm to calibrate the photo-detector. Under the aforementioned operating temperature and detection range, the software algorithm calculates the accuracy error between the reported optical power and the actual detected optical power to be within ±1 dB.

After calibrating the PD with this algorithm, the effects of dark current and noise can be eliminated, and the detection power range and detection accuracy can be improved in a wide temperature range.

The metal heatsink in the Optical Network Terminal (ONT) can couple the electromagnetic interference from high-speed microstrip lines and radiate electromagnetic waves into space. The electromagnetic radiation from the metal heatsink may have a serious impact on the wireless performance of the receiver. Therefore，in the design stage of ONT products，it is necessary to accurately simulate and evaluate the electromagnetic radiation of metal heatsinks.

This paper proposes a simulation method for electromagnetic radiation of metal heatsinks based on field-circuit co-simulation，which can comprehensively evaluate the resonance frequency and electromagnetic radiation intensity of the metal heatsink. Through the collaborative work of circuit simulation and electromagnetic field simulation，this method uses high-speed time-domain signals on microstrip lines as the excitation source of the model，making the simulation model closer to the actual situation. This paper verifies the rationality and accuracy of the simulation results obtained by this method through experiments.

The research results indicate that: Compared with the common simulation method，the electromagnetic radiation intensity of the heatsink obtained by this simulation method is in the order of －70 dBm，which is more in line with the actual level; The size of the metal heatsink is negatively correlated with its resonant frequency; By changing the size of the metal heatsink，the electromagnetic radiation intensity of the heatsink in the simulation is consistent with the measured Total Isotropic Sensitivity (TIS) metrics in the variation trend，and the correlation coefficient is above 0.94.

The proposed method is relatively accurate in simulating the electromagnetic radiation of metal heatsinks. During the design stage of ONT products，this method can be used to estimate the impact of electromagnetic radiation from metal heatsinks on wireless TIS indicators. This allows for targeted measures to be taken to suppress the electromagnetic radiation from metal heatsinks，thereby reducing debugging time during the prototype stage and improving research and development efficiency.

Mode-division multiplexing is a key technology to break through the capacity limit of single-mode fiber systems. In the few-mode fiber links, the pulse broadening caused by intermodal dispersion increases the complexity of multiple-input-multiple-output algorithms in the receiving terminal, which is one of the important reasons for limiting the implementation of long-haul transmission in mode-division multiplexing systems. The application of mode permutation methods can effectively compress pulse broadening and reduce mode-dependent loss, making it a mainstream technological approach for achieving long-distance, high-capacity mode-division multiplexing system transmission. However, there are numerous mode-permutation strategies with varying performance, and exploring mode-permutation strategies that are more suitable for long-distance, high-capacity mode-division multiplexing systems has become an urgent problem to be solved.

In this paper, we propose that the mode permutation strategy should match the mode group delay distribution of the fiber. Besides, we design a Mirror-Flipped Mode Permutation (MFMP) strategy based on few-mode fibers with symmetric mode delay. By modeling on a simulation platform, the transmission performance of mode-division multiplexing system based on the proposed MFMP is analyzed and verified through 6-mode multiplexing transmission experiments.

In this paper，based on the proposed MFMP strategy, 822-km 6-mode transmission over a mode-delay-symmetric few-mode fiber is experimentally demonstrated, carrying quadrature phase shift keying signals at 28 GBaud. At 616 km, compared to Cyclic Mode Permutation (CMP) and Cyclic Mode-Group Permutation (CMGP) methods, the signal pulse broadening is compressed by 37% and 6%, and the Q-factor is increased by 3.0 and 0.4 dB, respectively.

The proposed MFMP strategy can effectively compress pulse broadening, and improve the Q-factor of the system, while the advantages of this method over traditional two types of mode permutation strategies are verified via simulation and experiment. It should be noted that our work provides the theoretical basis and experimental guidance for achieving long-haul and high-capacity mode-division multiplexing transmissions based on mode permutation strategies.

The Extreme Learning Machine (ELM) neural network algorithm in the traditional indoor Visible Light Positioning (VLP) system suffers from unstable convergence and a tendency to get stuck in local optimal states, resulting in decreased positioning accuracy. In this paper, a Sparrow Search Algorithm (SSA)-ELM neural network algorithm was proposed by introducing the SSA to determine the initial weights and thresholds of the ELM neural network.

Firstly, the Received Signal Strength (RSS) and location information within the targeted area are collected as fingerprint data. Subsequently, the SSA-ELM neural network is trained to obtain a prediction model, and the test set data is input into this model to derive the positioning result for the location under test. Finally, simulation experiments and a testing platform are designed.

The simulation results show that at the four receiving heights of 0, 0.3, 0.6, and 0.9 m in the three-dimensional space model, the average errors are 1.73, 1.86, 2.18, and 3.47 cm, respectively. Compared with the Back Propagation (BP), SSA-BP, and ELM positioning algorithms, the positioning accuracy of the SSA-ELM algorithm is improved by 83.55%, 45.71%, and 26.26%, respectively, while the positioning time is reduced by 36.48%, 17.69%, and 6.61%, respectively. Experimental tests have shown that the average positioning error of the SSA-ELM neural network algorithm proposed in the article is 3.75 cm, representing a 16.38% improvement in positioning accuracy compared to the unoptimized ELM neural network.

The SSA-ELM neural network algorithm proposed in this paper has an obvious optimization effect on the ELM neural network, which can significantly reduce the positioning error of the system and reduce the positioning time of the system.

The dedicated power communication network is used for the dispatch, operation, and management of power systems, with its key characteristics being high reliability and real-time performance. At present, the typical relay distance of the power communication network is 40 km. In order to further improve its reliability, the number of optical amplifiers can be reduced by increasing the transmission distance between two optical nodes, which reduces the potential fault points as well.

In this work, a high-fidelity digital transmission scheme based on Delta-sigma modulation is proposed. At the transmitter, over-sampling, noise shaping, and two-level quantization are used to map the original high-order modulation signal into Quadrature Phase Shift Keying (QPSK) optical waveform, which improves the tolerance of the link noise and fiber nonlinearity. At the receiver, the link noise is removed by symbol decision, and then the out-of-band quantization noise is suppressed by a low-complexity low-pass filter for the demodulation of the original signal.

In the proof-of-concept experiment, after unrepeated transmission through 160 km standard single-mode fiber and coherent detection, 50 GBaud polarization-division-multiplexed QPSK optical waveform can achieve error-free transmission. At 9-times over-sampling, the recovered Signal-to-Noise Ratio (SNR) of original signal can reach 33.0 dB, which satisfies the 32.0 dB SNR threshold and supports high-fidelity transmission of 1 024-Quadrature Amplitude Modulation (1 024-QAM) signals with a single-wavelength bitrate of 111 Gbit/s.

Experimental results demonstrate that the Delta-sigma modulation scheme can achieve an improvement in SNR through bandwidth broadening, effectively extending transmission reach and performance. This provides a potential solution for the next generation of low-cost and high-reliability dedicated power communication networks.

Based on the micro-nano fiber coupler, this paper proposes a micro-nano fiber coupled ring mirror structure. Theoretical simulations and experimental studies are conducted to investigate the relationship between its spectrum and the polarization state of light within the ring mirror, the length of the micro-nano fiber in the tapered region, and the refractive index of the surrounding medium.

Theoretically, the basic principle of micro-nano fiber coupler is analyzed, and the optical field distribution of micro-nano fiber coupling and the spectral polarization of micro-nano fiber ring mirror are simulated. Experimentally, the spectral controllability is verified.

The results show that when the half-wave plate angle of the circular polarization controller is adjusted at －90～90°, the period and the peak wavelength remains constant. The extinction ratio changes regularly with the angle and reaches the minimum at 45°, as well as the maximum at 90°. As the length of the waist cone increases, the channel interval and the free spectral range of the ring mirror comb spectrum presents gradually decrease. As the surrounding refractive index varies in the range of 1.332 0 to 1.335 5, the output interference wavelength is red-shift with 18 350 nm / RIU sensitivity.

In summary, by adjusting the polarization state of the coupled light inside the micro-nano fiber coupler, the length of the micro-nano tapered region, and the refractive index surrounding the coupler structure, corresponding changes in the output spectrum can be achieved. This provides a new approach for constructing novel all-fiber devices. These devices can not only be used as wavelength selection and regulation components in lasers to optimize laser output characteristics, but also applied in sensing fields such as biomedicine for high-precision detection of liquid components.

With the rapid growth in demand for high-speed Passive Optical Networks (PONs), low-cost, low-noise Burst Mode Transimpedance Amplifiers (BM-TIAs) have become a key limiting factor. In this paper, a fast-settling, low-noise BM-TIA for 10 Gigabit Symmetrical Passive Optical Network (XGS-PON) is designed based on a low-cost 40 nm Complementary Metal Oxide Semiconductor (CMOS) process, which is compatible with three signal rates: 12.5, 10.0 and 2.5 Gbit/s.

To overcome the bandwidth and noise limitations of the CMOS process, the TIA achieves the required low noise through multi-stage amplification, an input inductor balancing network, and an increased power supply voltage. The gain adjustment is achieved by a combination of feedforward amplification and feedback resistance, enabling three gain levels and rate variations. To address burst-mode signals, fast direct current cancellation loop Automatic Offset Cancellation (AOC) is used to accurately eliminate the Direct Current (DC) input from the Avalanche Photodiode (APD). The charge-sharing techniques are employed to speed up the establishment of the offset cancellation loop and suppress the baseline drift by converting the AOC loop time constant.

The chip is designed and manufactured using a 40 nm CMOS process, with a die size of 945 μm×945 μm. The chip is tested with a commercial 10 Gbit/s APD in a Transistor Outline Can (TO-CAN) package. The test results show that the sensitivity of the chip at 12.5, 10.0 and 2.5 Gbit/s is －29.7，－33.0 and －37.6 dBm, respectively. The saturated input photocurrent can reach 2.5 mA at different data rates, and the chip achieves a large input dynamic range of 24.7, 28.2 and 32.8 dB for the three data rates. The static power consumption of the chip is 82.5 mW, and the AOC settling time in burst-mode is less than 23 ns across the entire input optical power range.

This chip, applied in XGS-PON, not only provides reference for the design of low-cost and low-noise BM-TIAs based on CMOS process but also has guiding significance for chip design in higher-speed PON scenarios.

With the growth of long distance oil and gas pipeline in our country, the optical communication network of oil and gas pipeline valve chamber plays an important role in the intelligent management of oil and gas pipeline. However, as an important basis for evaluating the optimization of optical communication network services, time delay faces challenges due to issues such as the existing optical communication network in oil and gas pipeline valve chambers and their relatively outdated technological systems. Problems with optical network Delay Measurement (DM) include low automation, long DM operation cycles, and significant impact on the normal operation of the carried services. Therefore, there is an urgent need for a new DM solution.

Based on the technology of Optical Transport Network (OTN) Optical Service Unit (OSU), a high precision delay calculation method is implemented by using the DM field in OSU frame structure.

Through experimental tests, under various business scenarios and granularity, the high-precision DM mechanism of optical communication network of oil and gas pipeline valve chamber designed in this paper can grasp its delay state in real time under the premise of ensuring DM accuracy, no manual field operation and normal business operation.

It is proved that the high-precision DM mechanism based on OTN OSU can provide strong support for the optimization and guarantee of service delay in optical communication networks within oil and gas pipeline valve chambers.

In order to study the temperature distribution and airflow requirements of high-speed optical modules, aiming to optimize the heat dissipation design and ensure the stable operation of optical modules.

The article adopts Flotherm simulation analysis to establish the numerical wind tunnel model of the optical module. The temperature distribution of each component during normal operation is obtained. The working air flow of the fan is also obtained when the system is stabilized, which are in line with the heat dissipation requirements in the specification of Multi-Source Agreement (MSA). In order to facilitate the testing and installation, the”L”type wind tunnel fixture is designed. The actual surface temperature of each chip and the actual working wind flow rate of the wind tunnel are tested experimentally to verify the accuracy of the simulation results.

The results show that in the case of the maximum temperature of 70 ℃, for the surface temperature of each chip, the difference between the simulation and the measured temperature is less than 2 ℃, with an error of less than 5%. For the airflow rate required for the heat dissipation of the entire optical module, the difference between the simulation and measured results is 0.2 Cubic Feet per Minute (CFM), with an error within 7%.

It can be shown that the simulation and measurement methods are feasible and the error is small, which provides an important reference value in the design and testing of heat dissipation of high-speed optical modules.

In the current rapidly advancing aerospace field, more stringent requirements are placed on the performance of fiber optic sensing demodulation systems. This study aims to address the wavelength hopping issue in the Modulated Grating Y-branch (MG-Y) laser during channel switching, so as to improve the wavelength stability of the laser and improve the resolution accuracy of the vibration signal in the fiber sensing demodulation system under high-speed scanning.

Through an in-depth investigation of the tuning characteristics of the MG-Y laser, we revealed the significant impact of switching frequency on the output wavelength. Then, we developed a wavelength hopping model for the MG-Y laser at different switching frequencies. Based on this model, we designed and implemented a wavelength hopping compensation system for the MG-Y laser. Additionally, we proposed a compensation method based on the internal channel current regulation. This method involves precise adjustment of the three-channel currents at the channel switching point to effectively compensate for wavelength errors and resolve the wavelength hopping issue.

The experimental results demonstrated a logarithmic relationship between switching frequency and wavelength deviation. By fitting the experimental data, we derived a relationship formula for wavelength deviation concerning switching frequency. Further experimental validation showed that by compensating for the three-channel currents at the channel switching point, wavelength errors were reduced to within 1 pm, indicating a significant improvement in performance.

Consequently, this compensation method led to a substantial reduction in wavelength deviation, enhancing the wavelength stability of the MG-Y laser. The article not only fills the technical gap in the field of optical fiber sensing demodulation systems under high-speed vibration environments in theory, but also directly contributes to enhancing the country's core competitiveness in the aerospace equipment sector, with large scientific and application value.