A fiber micro collimator with adjustable mode field diameter is proposed to address the problem of single output modefield diameter in traditional fiber collimators. The single mode fiber is combined with different 2-segment multimode fibers toform a "sandwich" structure. By adjusting the length of the middle section of the optical fiber, the mode field diameter can be adjusted arbitrarily, and a simulation experimental model has been constructed to prepare multiple fiber micro collimator sampleswith different sizes. The experimental results show that the structure can achieve adjustability of the output mode field diameter inthe range of 20~40 μm in the 1 550 nm wavelength band. The preparation method of this fiber collimator is simple, compact instructure, and can flexibly adjust the mode field diameter of the light field according to the needs to adapt to different applicationscenarios.

Passive harmonic mode-locking（HML）technology, as an effective method for rapidly increasing the output pulse repetitionfrequency of fiber lasers, significantly improves the output pulse repetition frequency of fiber lasers. This technologyachieves high-order harmonic output by integrating parameter adjustable saturable absorbers inside the laser, thereby significantlyincreasing the pulse repetition frequency. This paper summarizes several commonly used types of saturable absorbers in passiveHML fiber lasers, elaborates on their technical principles and advantages, and analyzes the latest technological progress in thisfield at home and abroad. Finally, this paper points out the future development trend of passive harmonic mode-locked fiberlasers.

Aming at the polarization-dependency issue of grating couplers, standard and improved polarization-insensitive gratingcouplers are successively designed on the thin film lithium niobate（TFLN） platform. During the design process, the particle swarmoptimization （PSO） algorithm is first improved, and then the improved PSO algorithm is used to optimize the grating parameters,including critical parameters such as grating period, duty cycle, etching depth, and incident angle. Finally, a polarization-insensitivechirped grating coupler is designed, and the effective refractive index is adjusted through linear chirp, significantly improvingthe coupling efficiency and polarization-insensitive bandwidth. The simulation results show that the coupling efficiency is thehighest at the wavelength of 1 544 nm, and the coupling losses of the transverse electric（TE） and transverse magnetic（TM） modesof the polarization-insensitive chirped grating coupler are -4.88 and -5.05 dB, respectively. The polarization-dependent loss（PDL）is less than 0.5 dB within the wavelength range of 1 500 nm to 1 580 nm, and the polarization-insensitive bandwidth is 45 nm.

Aiming at the issue of etalon effects in self-mode-locked lasers that affect laser output performance, a tunable-cavity-lengthNd:YVO4 self-mode-locked laser is designed. A single-end coated crystal is used to form an intra-cavity etalon for modeselection of the laser. The mode distribution of the spectrum and its corresponding timing diagram under different cavity lengthsare theoretically analyzed. The fluorescence spectrum temperature characteristics of the Nd:YVO4 crystal are experimentally verified,and its laser spectrum characteristics, the time trajectory of the mode-locked pulse sequence, and the spectrum distributionare analyzed respectively. The theoretical and simulation results show that when the ratio of the thickness of the etalon to the opticallength of the resonant cavity is a simple fraction of / , the longitudinal mode spacing can be modified to times the original,and a -order harmonic mode-locked laser is output. When the cavity length is 12.2 mm, and the ratio of the thickness of the etalonto the optical length of the resonant cavity is adjusted to 3/5, a high repetition frequency pulse laser output of 41.4 GHz isobtained. When the half-tube power in the vertical direction is 6 578 mW, the average output power exceeds 2.0 W, with the correspondingslope efficiency and light-to-light conversion rate of 31.2% and 30.7%, respectively.

To reduce the number and cost of optical network units in fiber-to-the-room systems, a half-duplex transceiver solution is proposed. This solution realizes the transceiver function of the communication system by connecting different circuit components through a mode switch. At the same time, the transceiver is used together with the WiFi wireless transceiver system to forma "fiber-wireless" half-duplex transceiver system architecture, enabling the transmission of optical signals and radio frequency signalsin the fiber-to-the-room system. The test results show that the designed mode switch and transceiver work smoothly together,and the signal eye diagram at the receiving end is clear when the data transmission rate is 5 Gb/s, verifying the feasibility of thehalf-duplex transceiver solution.

Optimizing the eye diagram performance of the transmitting optical signal and the sensitivity of the receiving signal ofthe optical module is the key to improving the transmission link budget of the high-speed optical network unit（ONU） optical module. A 25G ONU optical module with high speed performance is designed. First, the processing skills of high-speed signal lineimpedance discontinuity and the mechanism of the impact of printed circuit board（PCB） punching process design on the bandwidthperformance of transmission signal line are thoroughly studied. Then, the reference stack structure is adjusted and the via processmode is optimized. Finally, the high-speed signal simulation and analysis are carried out. The test results show that the designedmodule not only fully meets the performance requirements of the transmitting eye diagram and receiving sensitivity, but also has aperformance margin far exceeding the IEEE802.3ca protocol standard.

Given that the current research on amplitude-frequency effects is primarily limited to focusing on pulse top modulationfrom a time-domain perspective, which fails to comprehensively reveal its characteristics in the frequency domain or other dimensions, this study commences from a single-frequency laser phase modulation model, combining theoretical simulations and experimental verifications. It investigates the impact of single-order frequency loss, overall frequency loss, and modulation depthchanges on the excitation of amplitude-frequency effects, considering factors such as waveform intensity modulation, excited amplitude modulation signal strength, and spectrum distribution changes. The research results indicate that as the phase-modulatedlaser deviates further from the center position of the filter, the waveform intensity modulation increases accordingly, while the excited modulation signal strength exhibits a trend of first rising and then falling. Compared with single-order frequency loss, theamplitude-frequency effects excited by overall frequency loss mainly concentrate on the fundamental frequency and second harmonic. Additionally, an increase in modulation depth leads to an enhancement in spectral width and an intensification of the excited modulation signal strength.

In order to improve the sensitivity of fiber Bragg grating（FBG） type settlement sensors and reduce measurement deviations caused by tilt and temperature drift during the settlement process, an improved FBG sensor for structural settlement was designed. This sensor replaces the traditional cantilever structure with a straight pull structure to reduce measurement errors causedby settlement and tilt. By accurately calculating the ratio of wavelength offsets between two sets of FBGs, accurate measurementof settlement can be achieved. The test results show that the strain response of the sensor is 5.2 ×10-5 ε/N, and there is a linear relationship between wavelength shift and settlement within 0~50 mm. The average deviation is better than 0.69 mm, and the average sensitivity is 107.45 pm/mm. The fluctuation of the test results is below 1.0%, which is significantly better than the traditionalFBG sensor structure, verifying its better anti-interference ability.

In order to expand the applicable band range of the encoder, a multi-band Terahertz wave 4 ×2 encoder based ontwo-dimensional photonic crystals is designed. The performance of the encoder is thoroughly studied and analyzed using Rsoftsoftware combined with the plane wave expansion（PWE） method and the finite-difference time-domain（FDTD） method. The research results indicate that the output transmittance of the encoder varies with the lateral offset of the dielectric pillars at thewaveguide branch. After optimizing the design of the Terahertz wave 4×2 encoder, it can achieve high-contrast 4×2 encoding logic functions in the wavelength ranges of 103.80 ~108.57 μm, 109.19~111.12 μm, and 111.94~114.08 μm, with a contrast ratio ofno less than 9.30 dB and a maximum of 29.47 dB, and a response time is not more than 16.04 ps.

Aiming at the problems of large size, high insertion loss, and complex process of multi-mode interference （MMI） coupler, a 4×4 MMI coupler based on transverse electric （TE） mode is designed on the platform of X-cut thin-film lithium niobate（X-cut TFLN）. The length and width of the MMI region are simulated and optimized using the three-dimensional finite-difference time-domain （FDTD） method. Finally, a 4×4 MMI coupler is manufactured on the platform using inductive coupled plasma（ICP） etching method according to the optimization results. The simulation results show that the total insertion loss of the optimized 4×4 MMI coupler is 0.36 dB, and the actual measured insertion loss is 0.78 dB.

A tunable pulse width all solid state self-locking mode burst pulse laser is designed to address the problems of lowburst pulse power and complex device in current pulse lasers. The burst pulse generation mechanism of an intracavity Fabry-Perot（F-P） structure self-locking mode laser is theoretically analyzed, and a theoretical model of the self-locking mode burst pulselaser is established. The experimental results show that the pulse width of the burst pulse becomes wider as the length of the lasercrystal increases. When the length of the laser crystal is 5 mm and the pump power is 5.5 W, a burst pulse with a beam qualityfactor of 1.42 is obtained. The repetition frequency of the burst pulse is 13.64 GHz, the maximum average output power is 0.78 W,and the corresponding slope efficiency and optical-to-optical conversion efficiency are 9.57% and 8.73%, respectively.

In order to reduce the cost of Manchester decoding scheme in the traditional optical module industry and improve itscompatibility and reliability, a low-cost Manchester decoding scheme for optical modules is proposed. Firstly, the combined circuit of sampling resistor and opamp is used to preprocess the signal to solve the problem that the sampled signal amplitude is toosmall or saturated under different light intensity conditions and cannot be decoded. Secondly, the programmable counting array（PCA） inside the controller is used to capture the edge and calculate the duration of the signal to solve the problem of signal synchronization. In the process of synchronization judgment, the duration tolerance design is introduced, and the loss of data frameheader is optimized, which significantly improves the anti-interference ability of the decoding scheme. Finally, the programmablelogic unit（CLU） and timer inside the controller are used to extract the signal clock, and the signal is sent to the internal serial peripheral interface （SPI） module to complete the decoding, so as to improve the decoding efficiency and solve the problem of resource consumption and decoding compatibility of the controller. The test results show that the scheme can decode data accurately and stably, and meet the practical application requirements.

To address the issues of narrow gain bandwidth and complex structure in current fiber amplifiers, a 2 μm ultra-wideband PbS quantum dot fiber amplifier is fabricated. Firstly, PbS quantum dots with an average diameter of approximately 9.2 nmare synthesized through the organometrical approach and deposited on the surface of the optical fiber cone region by the opticaldeposition method. Subsequently, the PbS quantum dots coated in the conical region of the fiber are excited by a pumping lightsource to emit photons with the same wavelength as the signal light in the 2 μm band, thereby achieving the amplification of theoptical signal. The test results indicate that the optical fiber amplifier can attain a gain of 10.1 dB within the band of 1 900~2 100nm, and the gain bandwidth is approximately 200 nm.

To solve the problems of complex fiber structure and high manufacturing difficulty in existing methods of improvingthe birefringence of optical fibers, a hollow-core anti-resonant fiber with high birefringence characteristics is designed. Thecladding of the optical fiber is composed of a cladding ring, four half-tubes, and a pair of small circles nested inside the half-tubes.The finite element method is used to analyze the influence of the structural parameters of the fiber on its birefringence, transmission loss, confinement loss, and absorption loss. The simulation results show that changing the diameter of the half-tube has asmall effect on the birefringence. By adjusting the thickness difference between the horizontal and vertical half-tubes in the fiber,the fiber can achieve high birefringence, with a birefringence greater than 1.5×10-4 over a wide frequency range of 0.25 THz,while maintaining a transmission loss less than 10 dB/m.

Aiming at the challenge of enhancing throughput and maintaining high fairness in heterogeneous network access selection, a proximal policy optimization （PPO） algorithm based on reinforcement learning is proposed. This algorithm interacts withthe environment and samples data, aiming to maximize users' long-term throughput and satisfaction. It randomly simulates userlocations, collects user attribute data for model training, and acquires the optimal network access point allocation strategy. Thesimulation results show that compared with traditional algorithms, when the number of access users reaches the maximum, thePPO algorithm can increase throughput by 40% to 70%, while the average user satisfaction can still exceed 30%, and the userfairness index can reach 0.82.

In order to enhance the core capacity of optical fiber composite overhead ground wire（OPGW）, a 180 μm fine diameter single-mode optical fiber was developed. Research was conducted from two aspects: fiber profile structure design and optimized coating technology. The attenuation, stress-strain, and environmental performance of the 180 μm fine diameter OPGW optical cable were verified. The experimental results show that the attenuation of the entire 180 μm fine diameter fiber is stable during the cabling process. The tensile stress-strain performance and environmental performance of optical cables meet and exceedthe power industry standard requirements of OPGW cables, and have stable reliability and practical application feasibility.

To improve the transmission capacity of optical communication, the research goal of full-band large effective area optical fibers is proposed by analyzing the advantages and limitations of G.652.D and G.654.E optical fibers, as well as typical application scenarios of optical fiber transmission. Based on the hypothetical inference of fiber quality factor and compatibility, thepower exponent g of the gradient refractive index distribution of the full band large effective area fiber is analyzed and calculatedfrom formulas such as fiber refractive index distribution, normalized frequency of the fiber, and local aperture. The solution ideasfor the attenuation and bending performance of the new fiber are discussed, which expands new directions for the development ofnew fibers that can be compatible with the performance of G.652.D and G.654.E.

To improve the flatness of optical frequency combs （OFC） and the number of comb lines, a scheme for generating flatOFC based on dual light sources and polarization technology is proposed. This scheme first achieves eight flat comb lines otherthan the main carrier by adjusting the modulation index of the optical phase modulator（O-PM）. Then, it utilizes polarization technology to alter the peak power at the main carrier, resulting in a flat OFC. The simulation results indicate that when the O-PMmodulation index is set to 2.62, which corresponds to an amplitude E1 of 1.67 V for the radio frequency signal E1（t）, eight flat comblines excluding the central carrier can be generated. Under the conditions of a polarization controller angle φ of 29.8° and a linearpolarizer angle α of 45°, this scheme can produce a highly flat OFC with 10 comb lines, a flatness of 0.01 dB, and a side-modesuppression ratio of 6.15 dB. Compared to a single light source, the influence of linewidth changes in dual light sources on flatness is smaller.

With the gradual expansion of the communication network for oil and gas pipeline valve chambers and the increasingcomplexity of carrying business, the existing networking solutions urgently need to be upgraded and updated. Therefore, a communication network scheme for oil and gas pipeline valve chambers based on optical transport network optical service unit（OTNOSU） has been proposed. Firstly, the challenges faced in the networking process of oil and gas pipeline valve chamber communication networks were analyzed, and the current status of networking solutions was summarized. Then, the OTN OSU protocolwas introduced to optimize the optical network architecture for oil and gas pipeline valve chamber communication, and an industrial grade optical transmission customer front-end equipment （CPE） prototype was designed. Finally, conduct on-site environmental testing and verification of the plan. The test results show that the proposed scheme effectively improves network latencyand supports multiple business bearers. The power consumption of the CPE prototype is less than 40 W, which improves networkefficiency.