In view of the problem of short transmission distance in underwater visible light communication(VLC), this paper firstly analyzes the current situation of the development of VLC technology. According to the research ideas of arrayed design of transceiver, photonic design of detector, network design of communications node and diversification design of business types, the theoretical bottlenecks and key technologies that need to be further broken through in the future are proposed.

Aiming at the disadvantage of the high peak-to-average power ratio (PAPR) in asymmetrically clipped optical orthog-onal frequency division multiplexing (ACO-OFDM) systems in visible light communication, an improved discrete sine transform (DST) generalized precoding(DST-GP) method based on the GP method is proposed. In this method, the DST matrix, based on the GP method, is extended into an asymmetric matrix to obtain the better PAPR reduction performance, and then the calculation of the extended part is optimized by the GP method. The simulation results and analyses show that the proposed DST-GP method, compared with the original DST method, has a better PAPR reduction performance with the low cost and can achieve a better compromise between the PAPR reduction performance and the computational complexity.

Aiming at the problem that the optical cable is often damaged by engineering construction, a low-cost vibration early warning and positioning system for monitoring the engineering construction along the optical cable line is proposed. Multiple vibration early warning modules based on single axis Sagnac interferometer are used to monitor the vibration of multiple optical cables in real time, and the position of the vibration point is determined by the phase sensitive time domain reflectometer(Φ- OTDR) module. Experiments show that the system has low cost and good early warning effect.

The prediction of transmission quality(QoT) is becoming more and more important in optical networks. Machine learning has become an important means to realize the prediction of QoT in optical networks in the future. This paper presents a new technology of QoT prediction based on machine learning classifier. Through the data generated by the transfer equation, it can be used for classifier training and performance testing. The performance of K-Nearest Neighbor(KNN), logistic regression(LR) and support vector machine(SVM) are verified by simulation. The simulation results show that, compared with the traditional QoT estimation method, the machine learning based on method can effectively reduce the computational complexity and provide quite high prediction accuracy. This method is a new QoT estimation scheme with broad application prospects.

With the wide application of repeatless optical fiber transmission system, its complexity and cost become factors to be considered. A simplified transmission system without relay is proposed, in which Stokes vector direct detection(SV-DD) is used instead of coherent receiver. The performance of 128 Gb/s 16QAM signal transmission system with single wave and three wave multiplexing is simulated. The simulation results show that the system can demodulate the signal well, and the performance of forward and backward Raman pump is better when the power is 0.6 W and 0.8 W, and it can withstand the residual dispersion of about 400 ps · nm-1.

In order to achieve low-cost, long-distance and large capacity data transmission scenarios, a fiber-optic bidirectional optical fiber transmission system without multiple in multiple out(MIMO) mode division multiplexing(MDM) adjustment bandwidth is designed. Using six modes to select the photon lantern(PL) as the mode multiplexer and demultiplexer respectively, LP21 and LP01 are selected two non degenerate modes as forward transmission channels, LP02 mode and LP01 mode as reverse transmission channels. At the same time, two degenerate modes of LP11 are selected as forward and reverse transmission channels through optical switch control, thus realizing 12.5 km bidirectional bandwidth adjustable transmission. The system provides a solution for low-cost, long-distance and large capacity data transmission scenarios.

In order to improve the output energy stability of fiber front-end system, an power automatic control optical fiber amplifier is designed with closed-loop feedback. The amplifier output signal splitting beem is used as a feedback signal. After photoelecrtic conversion and sampling by the high-speed peak hold circuit, the proportional integral differential(PID) algorithm is used to achieve an accurate control of the amplifier pump current. The experimental results show that under the condition that the energy of the injected large light pulse fluctuates, the P-V value of the fiber amplifier output energy is less than 8% and the root mean square(RMS) value is less than 1% in two hours. The stability of the output energy has been effectively improved.

In the existing microwave photons system, low-cost, high-performance and arrayed electro-optic conversion components become one of the technical bottlenecks that restrict the application of microwave photons. A four channel direct modulation electro-optic conversion module based on photoelectric hybrid packaging technology is designed. The module integrates a variety of functional chips and realizes the function of converting four channels of 2~18 GHz radio frequency signals into optical signals. The test results show that the module achieves 18 GHz electro-optic modulation bandwidth, 9 dB insertion loss, less than 30 dB noise coefficient and more than 10 dBm input third order intersection3(IIP3). While ensuring high-performance electro-optical conversion, it can effectively reduce package size, reduce system weight and improve system integration, which is conducive to the array application of microwave photons technology.

Aiming at the problems of traditional manual alignment time-consuming, low yield and poor consistency of transparent optical communication devices, the edge model and coupling model of transparent optical communication devices are established by studying the automatic assembly requirements of transparent optical communication devices. The influence of edge error on coupling loss is analyzed by overlapping integral method, and an edge detection of transparent optical communication devices based on digital image is designed programme. The experimental results show that the scheme can detect the edge of transparent optical communication device in complex background well, and can meet the accuracy requirements of alignment deviation. It provides support for the research of intelligent alignment system of automatic assembly of transparent optical communication device, and has important engineering application significance.

In the dense wavelength division multiplexing optical fiber communication system, the gain flatness of the Raman fiber amplifier is low and the bandwidth is narrow. Starting with the coupling equation of the front-end multi-pumped Raman amplifier, using the Runge Kutta method and the tellurium based fiber as the gain fiber of the multi-pumped Raman amplifier, the number of pumps, pump power and light in the multi wavelength pumping are analyzed by Matlab. Finally, the simulation of 100 nm transmission bandwidth is realized by using six pumps and setting pump power reasonably.

The optical node wavelength multiplexer / demultiplexer and optical switch matrix in optical transport network(OTN) can map the logical topology of any structure on the physical topology, the unreasonable mapping scheme will consume additional port resources. A logic topology optimization mapping algorithm based on reinforcement learning(RL) is proposed. The preprocessed topological state and logical channel data are used to train the RL model, so as to allocate the global wavelength resources to the logical channel, and finally achieve the purpose of resource optimization. Simulation results show that the proposed algorithm can effectively reduce the resource consumption in the process of logical topology mapping, thus minimizing the cost of network deployment.

In view of the existing multi-pulse position modulation (MPPM) schemes in deep space optical communication, the constellation points have a low utilization rate and the location selection needs to be optimized. By comparing the average hamming distance between constellation points of different MPPM schemes and the probability of signal pulses in adjacent time slots, an MPPM constellation mapping scheme with stronger anti-delay jitter capability is proposed. Based on the existing serial concatenated pulse position modulation(SCPPM) decoding algorithm, a SCMPPM iterative decoding algorithm is derived. Simulation results show that the performance of this scheme is better than the existing schemes. It not only enables the communication system to maintain a high information transmission rate, but also reduces the probability of inter-symbol interference(ISI) and inter-slot interference(ITI) between MPPM constellations, thereby improving the anti-delay jitter capability of communication system.

Aiming at the problem of high peak to average power ratio in optical orthogonal frequency division multiplexing (OOFDM) system, the coordinate rotation digital computer(CORDIC) algorithm is used for nonlinear compression, and field programmable gate array(FPGA) is used for hardware implementation. The signal captured from FPGA is analyzed on Matlab, the peak to average power ratio and bit error rate of communication system are calculated, and compared with the traditional OOFDM system. Experimental results show that the algorithm can effectively suppress the peak to average power ratio of the signal and ensure the communication quality.