A method and apparatus for dispersion programmed frequency conversion based on silicon core fiber

Abstract

The application discloses a dispersion programming frequency conversion method and device based on a silicon core optical fiber. The method acquires dispersion parameters, nonlinear parameters and loss parameters of the silicon core optical fiber under different taper diameters, cladding structures, pump wavelengths and temperature conditions, constructs a dispersion-nonlinear-loss joint constraint model according to wavelengths of to-be-converted signal light and target idler light, determines a taper waist region diameter distribution, a pump wavelength, a double-pump interval and an input polarization state, so that pump light, signal light and idler light satisfy quasi-phase matching conditions, and adjusts pump power and fiber temperature according to output spectrum feedback, thereby realizing selective frequency conversion output of target wavelengths.

Description

Technical Field

[0001] This invention relates to the fields of optical fiber communication, nonlinear optics, and silicon core optical fiber devices, and specifically to a dispersion programming frequency conversion method and apparatus based on silicon core optical fiber. Background Technology

[0002] All-optical wavelength conversion technology can shift the wavelength of signal light without photoelectric conversion, and has application needs in scenarios such as wavelength division multiplexing communication, all-optical switching, optical parametric amplification, extended-band light sources, and spectral detection. Four-wave mixing based on third-order nonlinear effects is an important method for achieving all-optical wavelength conversion. Its basic process involves inputting pump light and signal light into a nonlinear medium, and generating new idler light through energy conservation and phase matching. The conversion efficiency, target wavelength selectivity, and output stability of this type of technology are typically affected by factors such as the nonlinear coefficient, group velocity dispersion, transmission loss, pump power, effective operating length, and polarization state.

[0003] Existing fiber-based wavelength conversion schemes typically employ highly nonlinear fibers, photonic crystal fibers, or other special fibers as the nonlinear medium. Highly nonlinear fibers improve conversion efficiency by increasing the nonlinear coefficient and reducing dispersion values ​​near specific wavelength bands, while photonic crystal fibers adjust waveguide dispersion through air hole size, aperture spacing, and refractive index distribution. These schemes can achieve four-wave mixing and frequency conversion within a certain wavelength band, but their dispersion adjustment mainly relies on a fixed cross-sectional structure. After fiber fabrication, the dispersion distribution along the propagation direction is often difficult to reconstruct according to different target wavelength conversion tasks. When the signal light and the target idler light span a wide wavelength band, the fixed dispersion structure is prone to problems such as limited phase matching range, enhancement of non-target sidebands, increased pump threshold, or decreased conversion efficiency.

[0004] On-chip silicon waveguides leverage the high third-order nonlinearity and strong waveguide confinement capabilities of silicon to achieve all-optical wavelength conversion. This type of solution is suitable for integrating photonic chips, but when connected to external fiber optic communication systems, it typically requires end-face coupling, grating coupling, or lens coupling structures. Coupling loss and packaging precision can affect system efficiency. The effective operating length of the on-chip waveguide is also limited by factors such as chip size, propagation loss, thermal effects, and nonlinear absorption. For frequency converters requiring direct embedding of fiber optic links, on-chip solutions often necessitate additional fiber-to-chip coupling packaging, resulting in a more complex system structure.

[0005] Silicon-core optical fiber incorporates silicon material in its core and uses glass as the cladding, combining the high nonlinearity of silicon with the ease of transmission, coupling, and post-processing inherent in optical fiber. By tapering the silicon core, the optical field constraint can be enhanced and waveguide dispersion adjusted by reducing the silicon core diameter, making it suitable for nonlinear frequency conversion, optical parametric amplification, and extended-band light source generation. However, existing silicon-core fiber frequency conversion schemes often focus on verifying that a specific tapered structure can generate four-wave mixing or wavelength conversion. They typically lack a reverse-solvable dispersion design method for the wavelengths of the signal to be converted and the target idler wavelength, and they do not incorporate group velocity dispersion, nonlinear coefficients, and transmission loss into a unified constraint model.

[0006] Furthermore, the actual tapering process of silicon core optical fibers involves diameter errors, local loss variations, temperature drift, and polarization state fluctuations. If only a pre-defined structure is used for open-loop operation, the output idler light may exhibit power instability, center wavelength shift, or non-target sideband enhancement. Therefore, existing technologies still require a frequency conversion method and device capable of designing the axial dispersion profile of the silicon core optical fiber according to the target wavelength conversion task, and capable of combining output spectral feedback to correct pump power, fiber temperature, and input polarization state, in order to improve the designability, selectivity, and stability of the target idler light output. Summary of the Invention

[0007] To address the shortcomings of existing fiber-based or on-chip nonlinear wavelength conversion schemes, such as insufficient flexibility in target wavelength conversion, difficulty in designing phase-matching conditions according to the target wavelength, limited adaptability of fixed dispersion structures to cross-band frequency conversion, and instability of output idler light due to tapering processing errors and temperature drift, this invention provides a dispersion-programmed frequency conversion method and device based on silicon core optical fiber. This method uses the wavelength of the signal light to be converted and the wavelength of the target idler light as inputs to obtain the dispersion, nonlinear, and loss parameters of the silicon core optical fiber under different tapering diameters, cladding structures, pump wavelengths, and temperatures. A joint constraint model of dispersion, nonlinearity, and loss is constructed, and based on this, the axial diameter distribution, pump wavelength, double-pump spacing, and input polarization state of the tapered waist region of the silicon core optical fiber are determined, ensuring that the signal light, pump light, and idler light satisfy quasi-phase-matching conditions within the effective propagation length of the silicon core optical fiber. Furthermore, by acquiring the output spectrum, feedback corrections are made to the pump power, fiber temperature, and input polarization state, thereby obtaining selective output of the target idler light.

[0008] One aspect of this invention provides a dispersion-programmed frequency conversion method based on silicon-core optical fiber. The method includes acquiring the wavelength of the signal light to be converted and the wavelength of the target idler light. The signal light to be converted can be a continuous wave signal light or an optical signal carrying communication modulation information. The target idler light is the target frequency-converted light generated by third-order nonlinear interactions in the silicon-core optical fiber and is the desired output. The wavelengths of the signal light to be converted and the target idler light can be located in the communication O-band, E-band, S-band, C-band, L-band, U-band, or extended bands around 2 μm. By using the wavelengths of the signal light to be converted and the target idler light as initial inputs, this invention enables parameter design based on the target wavelength conversion task, rather than passively testing its convertible band after the fixed-structure device is fabricated.

[0009] This method further includes obtaining dispersion parameters, nonlinear parameters, and loss parameters of silicon-core optical fibers under different tapered diameters, cladding structures, pump wavelengths, and temperatures. The silicon-core optical fiber comprises a silicon core and a cladding. The silicon core can be monocrystalline silicon, polycrystalline silicon, recrystallized silicon, or low-defect silicon material, and the cladding can be quartz glass, borosilicate glass, fluorine-doped glass, or other transparent glass materials suitable for forming the fiber cladding. The dispersion parameters may include effective refractive index, propagation constant, group velocity dispersion, and higher-order dispersion. The nonlinear parameters may include effective mode area and nonlinear coefficient. The loss parameters may include transmission loss, insertion loss, and additional loss introduced by the tapered transition region. By calculating or testing the modes of silicon-core optical fibers with different tapered diameters and cladding structures, the propagation constant, group velocity dispersion, and nonlinear coefficient at different axial positions can be obtained, thereby forming a silicon-core optical fiber parameter database for subsequent model solving.

[0010] During parameter acquisition, the silicon core fiber can be divided into multiple axial sampling positions along the light propagation direction. At each axial sampling position, the silicon core diameter, cladding equivalent refractive index, local temperature, effective refractive index, propagation constant, group velocity dispersion, effective mode area, nonlinear coefficient, and transmission loss are recorded or calculated. Since the waist diameter of the silicon core fiber can vary along the light propagation direction after tapering, the dispersion and nonlinear parameters corresponding to each axial sampling position can also be different. In this way, subsequent calculations no longer treat the silicon core fiber as a uniform nonlinear medium with a single average diameter, but rather as a frequency conversion medium with axial dispersion and axial nonlinear distributions.

[0011] This method further constructs a joint constraint model of dispersion-nonlinearity-loss based on the wavelength of the signal light to be converted and the wavelength of the target idler light. The joint constraint model includes energy conservation constraints, axial phase mismatch constraints, loss constraints, and target idler light selectivity constraints. In single-pump frequency conversion mode, the pump light, signal light, and idler light satisfy 2ωp=ωs+ωi, where ωp is the angular frequency of the pump light, ωs is the angular frequency of the signal light, and ωi is the angular frequency of the idler light. In dual-pump frequency conversion mode, the first pump light, second pump light, signal light, and idler light satisfy ωp1+ωp2=ωs+ωi, where ωp1 and ωp2 are the angular frequencies of the first and second pump lights, respectively. Through the energy conservation relationship, candidate pump wavelengths or candidate dual-pump combinations can be generated based on the wavelength of the signal light to be converted and the wavelength of the target idler light.

[0012] This method also determines whether candidate structures and candidate pump parameters can achieve quasi-phase matching based on the axial phase mismatch. In single-pump frequency conversion mode, the axial phase mismatch can be expressed as Δβ(z)=2βp(z)-βs(z)-βi(z)+2γ(z)Pp(z). Where z is the axial position of the silicon core fiber, βp(z), βs(z), and βi(z) are the propagation constants of the pump light, signal light, and idler light at the axial position z, respectively, γ(z) is the nonlinear coefficient at the axial position z, and Pp(z) is the power of the pump light at the axial position z. In dual-pump frequency conversion mode, the axial phase mismatch can be expressed as Δβ(z)=βp1(z)+βp2(z)-βs(z)-βi(z)+ΔβNL(z). The axial phase mismatch is calculated by ΔβNL(z), where ΔβNL(z) is a nonlinear phase shift correction term determined by the nonlinear coefficient, the power of the first pump light, the power of the second pump light, and the polarization coupling coefficient. βp1(z) and βp2(z) are the propagation constants of the first pump light and the second pump light at axial position z, respectively. Pp1(z) and Pp2(z) are the powers of the first pump light and the second pump light at axial position z, respectively.

[0013] Preferably, the quasi-phase matching condition in this method is not limited to zero phase mismatch at a fixed cross-section, but rather ensures that the axial phase mismatch integral over the effective propagation length of the silicon core fiber is within a preset tolerance range. Specifically, it can make | _0^L Δβ(z)dz-2πm|≤ε, where L is the effective propagation length, m is the quasi-phase matching order, and ε is the phase tolerance. Since the diameter distribution, temperature distribution, pump power distribution, and input polarization state of the tapered waist region can all affect Δβ(z), this invention, through the joint design of these variables, ensures that the cumulative phase mismatch corresponding to the target idler light meets the quasi-phase matching condition, and causes the non-target idler light to deviate from this condition, thereby achieving selective enhancement of the target idler light.

[0014] Preferably, loss is also considered as a constraint in the joint constraint model. After tapering the silicon core fiber, the reduced effective mode area can improve the nonlinear coefficient, but an excessively small silicon core diameter may increase scattering loss, mode leakage, or additional loss in the transition region. Therefore, this invention considers both the nonlinear coefficient and loss parameters when determining the diameter distribution of the taper waist region. When the loss can be approximated as uniform loss, the effective operating length can be expressed as Leff = [1 - exp(-αL)] / α, where α is the transmission loss coefficient. When the loss varies along the axial position, the equivalent operating length can be calculated based on the axial loss distribution. Through loss constraints, this invention avoids the situation where simply reducing the taper diameter to increase nonlinearity leads to excessively high total loss.

[0015] Preferably, when determining the axial diameter distribution of the tapered waist region of a silicon core optical fiber, a structure satisfying the target conditions can be selected from multiple candidate structures based on a joint constraint model. The axial diameter distribution can be segmented stepped, continuously tapered, periodically modulated, or a combination thereof. A segmented stepped axial diameter distribution includes at least two dispersion control segments with different silicon core diameters, each corresponding to different group velocity dispersion, nonlinear coefficients, and transmission losses. A continuously tapered axial diameter distribution causes the silicon core diameter to change continuously along the light propagation direction to reduce mode perturbations between adjacent segments. A periodically modulated axial diameter distribution causes the silicon core diameter to change periodically around the average diameter along the light propagation direction to provide an equivalent quasi-phase matching condition for the target idler light. A composite axial diameter distribution can combine segmented stepped regions, continuously tapered regions, periodically modulated regions, and flat high-nonlinear regions to balance input and output losses, nonlinear enhancement, and phase mismatch compensation.

[0016] Preferably, when determining the pump parameters, either a single-pump mode or a dual-pump mode can be used. The single-pump mode has a relatively simple structure and is suitable for wavelength conversion tasks where the signal light and target idler light revolve around the pump light frequency distribution. The dual-pump mode, by having both the first and second pump lights participate in nonlinear frequency conversion, can increase the selection space for pump wavelength combinations and is suitable for frequency conversion tasks with a larger wavelength span. The dual-pump interval can be determined based on the target idler light wavelength, the wavelength of the signal light to be converted, and the axial dispersion distribution of the silicon fiber, and can also be fine-tuned during the output spectrum feedback process.

[0017] Preferably, this method further inputs the signal light and pump light into the silicon core fiber, so that the signal light, pump light, and idler light satisfy the quasi-phase matching condition in the silicon core fiber and generate the target idler light. The signal light and pump light can be coupled into the silicon core fiber through a wavelength division multiplexer, a broadband coupler, a polarization combiner, or other combining devices. Before input, the polarization state of the signal light and pump light can be adjusted by a polarization controller to ensure that their effective nonlinear coupling state in the silicon core fiber meets preset requirements. For polarization-sensitive silicon core fiber structures, a polarization scanning method can be used to find the input polarization state with higher target idler light power or higher sidemode suppression ratio.

[0018] This method also includes acquiring the output spectrum and correcting at least one of the pump power, fiber temperature, and input polarization state based on the output spectrum feedback. The output spectrum may include residual signal light, residual pump light, target idler light, and non-target sidebands. The center wavelength of the target idler light, the power of the target idler light, the power of the residual pump light, the power of the residual signal light, and the power of the non-target sidebands can be identified by the output spectrum, and the conversion efficiency, sidemode suppression ratio, and target idler light wavelength offset can be calculated further. When the target idler light power is lower than a preset value, the pump power or the input polarization state can be adjusted; when the center wavelength of the target idler light deviates from the preset wavelength, the pump wavelength, the dual-pump spacing, or the fiber temperature can be adjusted; when the power of the non-target sidebands exceeds a preset threshold, the fiber temperature, the dual-pump spacing, or the input polarization state can be adjusted to make the non-target sidebands deviate from the phase matching condition.

[0019] Another aspect of the present invention provides a dispersion-programmable frequency converter based on silicon-core optical fiber. The device includes a signal light input unit, a pump light source unit, a polarization control unit, a beam combining unit, a silicon-core optical fiber dispersion-programmable frequency converter, a temperature control unit, an output filtering unit, a spectral monitoring unit, and a feedback control unit. The signal light input unit is used to input the signal light to be converted; the pump light source unit is used to output single-pump or dual-pump pump light; the polarization control unit is used to adjust the input polarization state of at least one of the signal light and the pump light; the beam combining unit is used to couple the signal light and the pump light to the silicon-core optical fiber dispersion-programmable frequency converter; the silicon-core optical fiber dispersion-programmable frequency converter has a tapered waist region, and the tapered waist region has a silicon core diameter distribution that varies along the light propagation direction; the temperature control unit is used to adjust the temperature of the silicon-core optical fiber dispersion-programmable frequency converter; the output filtering unit is used to separate the target idler light; the spectral monitoring unit is used to acquire the output spectrum; and the feedback control unit is used to adjust at least one of the pump light source unit, the temperature control unit, and the input polarization state according to the output spectrum.

[0020] Preferably, the silicon core fiber dispersion-programmable frequency converter unit may include an input transition cone region, a tapered waist region, and an output transition cone region. The input transition cone region is used to transition the input-side silicon core diameter to the starting diameter of the tapered waist region, and the output transition cone region is used to transition the end diameter of the tapered waist region to the output-side silicon core diameter. The tapered waist region is used to provide the main nonlinear interaction and phase mismatch adjustment. The tapered waist region may include multiple dispersion control segments, and may also have a continuously varying diameter distribution or a periodically varying diameter distribution. Each dispersion control segment or each axial sampling position corresponds to a set of effective refractive index, group velocity dispersion, nonlinear coefficient, and transmission loss parameters, so that the tapered waist region forms a dispersion distribution in the optical propagation direction for adjusting the cumulative phase mismatch.

[0021] Preferably, the pump source unit may include a first pump source and a second pump source, and a dual-pump interval control module may be provided between the first pump source and the second pump source. The temperature control unit may be an integral temperature control unit or a zoned temperature control unit. When a zoned temperature control unit is used, multiple temperature control zones may be set along the axial direction of the silicon core fiber dispersion programming frequency converter unit to adjust the temperature of the input transition zone, the main nonlinear action zone, and the output compensation zone respectively. The device may also include a polarization control unit, which is disposed on at least one of the signal light input path and the pump light input path, and is used to adjust the effective nonlinear coupling state of the signal light and the pump light in the silicon core fiber.

[0022] Preferably, the feedback control unit can store a silicon core fiber parameter database and control parameters corresponding to the target wavelength conversion task. The silicon core fiber parameter database includes dispersion parameters, nonlinear parameters, and loss parameters under different tapered diameters, cladding structures, pump wavelengths, and temperature conditions. The feedback control unit can call or solve for the axial diameter distribution of the tapered waist region, pump wavelength, dual-pump spacing, pump power, input polarization state, and temperature control parameters based on the wavelength of the signal light to be converted and the target idler light wavelength. It can also correct the control parameters when the target idler light power, center wavelength, or side-mode suppression ratio does not meet preset conditions, based on feedback from the spectral monitoring unit.

[0023] In summary, this invention first establishes a parameter database for silicon-core optical fiber under different tapered structures and operating conditions, using the target wavelength conversion task as input. Then, it determines the axial diameter distribution and pump control parameters of the tapered waist region through a dispersion-nonlinearity-loss joint constraint model, ensuring that the signal light, pump light, and target idler light satisfy the quasi-phase matching condition within the effective propagation length. Subsequently, the signal light and pump light are input into the silicon-core optical fiber dispersion-programmed frequency conversion unit to generate the target idler light, and the pump power, temperature, and polarization state are adjusted based on the output spectrum feedback. Through this process, the silicon-core optical fiber is no longer merely a fixed-structure nonlinear transmission medium, but rather a frequency conversion medium capable of axial dispersion design and closed-loop adjustment according to the target wavelength conversion task.

[0024] The advantages of this invention over existing technologies are as follows: By using the wavelength of the signal light to be converted and the wavelength of the target idler light as inputs, and combining the dispersion parameters, nonlinear parameters, and loss parameters of silicon core optical fiber under different tapered diameters, cladding structures, pump wavelengths, and temperature conditions, a joint constraint model of dispersion-nonlinearity-loss is established. Based on this model, the axial diameter distribution, pump parameters, input polarization state, and temperature adjustment amount in the tapered waist region are determined, ensuring that the phase mismatch accumulation corresponding to the target idler light is within a preset tolerance range, thereby improving the designability of the target wavelength conversion. Simultaneously, the highly nonlinear core and fiber morphology propagation structure of the silicon core optical fiber facilitate the accumulation of nonlinear interactions over a longer effective operating length, reducing dependence on on-chip fiber coupling packaging. Furthermore, by correcting the pump power, fiber temperature, and input polarization state through output spectral feedback, the effects of tapered processing errors, temperature drift, and polarization fluctuations on phase matching conditions can be compensated, non-target sideband enhancement can be suppressed, and the selectivity and stability of the target idler light output can be improved. This makes the method and device suitable for designable all-optical wavelength conversion, optical parametric amplification, and extended band light source generation within the extended band of communication. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is the overall flowchart of the dispersion programming frequency conversion method based on silicon core optical fiber of the present invention.

[0027] Figure 2 This is a structural block diagram of the dispersion-programmable frequency converter based on silicon core optical fiber according to the present invention.

[0028] Figure 3 This is a schematic diagram of the axial structure of the silicon core optical fiber dispersion programming frequency conversion unit of the present invention.

[0029] Figure 4 This is a schematic diagram of the diameter distribution of the segmented stepped tapered waist region of the present invention.

[0030] Figure 5 This is a schematic diagram of the diameter distribution of the waist region of the continuously tapered conical shape of the present invention.

[0031] Figure 6 This is a schematic diagram of the periodic modulation type tapered waist region quasi-phase matching of the present invention.

[0032] Figure 7 This is a flowchart of the calculation of the dispersion-nonlinearity-loss joint constraint model of the present invention.

[0033] Figure 8 This is the output spectrum feedback control flowchart of the present invention.

[0034] Figure 9 This is a schematic diagram of the application scenarios and output spectrum of the present invention for extended band optical communication. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0036] It should be understood that in the description of this invention, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0037] In this invention, the wavelength of the signal light to be converted is denoted as λs, the wavelength of the target idler light is denoted as λi, and the wavelength of the pump light is denoted as λp. In dual-pump mode, the wavelengths of the first and second pump lights are denoted as λp1 and λp2, respectively. The corresponding angular frequencies are denoted as ωs, ωi, ωp, ωp1, and ωp2, respectively. The axial position of the silicon core fiber is denoted as z, the diameter distribution of the silicon core in the tapered waist region is denoted as d(z), the propagation constant is denoted as β(ω,z), the nonlinear coefficient is denoted as γ(z), the transmission loss is denoted as α(z), and the axial phase mismatch is denoted as Δβ(z). Example 1

[0038] This embodiment illustrates the complete implementation process of the method of the present invention. See also... Figure 1 This embodiment provides a dispersion programming frequency conversion method based on silicon core optical fiber. The method takes the wavelength of the signal light to be converted and the wavelength of the target idler light as input, and achieves selective output of the target idler light through silicon core optical fiber parameter acquisition, joint constraint modeling, determination of axial diameter distribution in the tapered waist region, configuration of pump and polarization parameters, nonlinear frequency conversion, and output spectrum feedback correction.

[0039] First, the target wavelength conversion task is obtained. Specifically, the wavelength of the signal light to be converted, λs, and the target idler light wavelength, λi, are set, and the allowable pump wavelength range, pump power range, temperature regulation range, and target output specifications are determined. Target output specifications may include the allowable deviation of the target idler light center wavelength, conversion efficiency, side-mode rejection ratio, and upper limit of non-target sideband power. For example, in one embodiment for communication band relocation, λs can be located in the range of 1530 nm to 1565 nm, and λi can be located in the range of 1565 nm to 1625 nm; in another embodiment for cross-band frequency conversion, λs can be located in the range of 1260 nm to 1360 nm, and λi can be located in the range of 1530 nm to 1565 nm; in yet another embodiment for extended band light sources, λs can be located in the range of 1530 nm to 1565 nm, and λi can be located in the range of 1900 nm to 2400 nm. The above wavelength ranges are used to illustrate optional application scenarios and do not constitute a limitation of the present invention.

[0040] Next, a database of silicon core fiber parameters is established. The silicon core fiber 200 comprises a silicon core 201 and a cladding 202. The silicon core 201 can be made of monocrystalline silicon, polycrystalline silicon, recrystallized silicon, or low-defect silicon, while the cladding 202 can be made of quartz glass, borosilicate glass, fluorine-doped glass, or other low-loss transparent glass materials. For the silicon core fiber to be used, its effective refractive index, propagation constant, group velocity dispersion, higher-order dispersion, effective mode area, nonlinear coefficient, and transmission loss are obtained under different silicon core diameters, different cladding structures, different pump wavelengths, and different temperatures. These parameters can be obtained through mode solving, spectral measurement, loss testing, white light interferometry, or four-wave mixing response inversion, or determined through a combination of simulation calculations and experimental calibration.

[0041] When establishing the parameter database, the silicon core fiber is divided into multiple axial sampling positions along the light propagation direction. For each axial sampling position zm, the silicon core diameter d(zm), local temperature T(zm), effective refractive index neff(λ,zm), propagation constant β(ω,zm), group velocity dispersion β2(zm), effective mode area Aeff(zm), nonlinear coefficient γ(zm), and transmission loss α(zm) at that position are recorded or calculated. Since the silicon core diameter within the tapered waist region 205 can vary along the axial direction, each set of parameters in the parameter database corresponds to a local structural state. In this way, the silicon core fiber can be regarded as a nonlinear frequency conversion medium with axially varying dispersion and axially varying nonlinear coefficient in subsequent calculations.

[0042] Subsequently, a joint constraint model of dispersion, nonlinearity, and loss is constructed based on the target wavelength conversion task. For the single-pump mode, candidate pump wavelengths are determined according to the energy conservation relation 2ωp=ωs+ωi. For the dual-pump mode, candidate combinations of the first and second pump lights are determined according to the energy conservation relation ωp1+ωp2=ωs+ωi. After determining the candidate pump wavelengths, the propagation constants of the signal light, pump light, and target idler light at different axial positions are calculated based on the parameter database, and the axial phase mismatch is further calculated.

[0043] In single-pump mode, the axial phase mismatch can be expressed as: Δβ(z)=2βp(z)-βs(z)-βi(z)+2γ(z)Pp(z) Where βp(z), βs(z), and βi(z) are the propagation constants of the pump light, signal light, and target idler light at axial position z, respectively, and Pp(z) is the power of the pump light at axial position z. Since the pump light is affected by transmission loss and nonlinear interaction during propagation, Pp(z) can be calculated based on the input pump power and axial loss distribution.

[0044] In dual-pump mode, the axial phase mismatch can be expressed as: Δβ(z)=βp1(z)+βp2(z)-βs(z)-βi(z)+ΔβNL(z) Where βp1(z) and βp2(z) are the propagation constants of the first pump light and the second pump light at axial position z, respectively; Pp1(z) and Pp2(z) are the powers of the first pump light and the second pump light at axial position z, respectively; and ΔβNL(z) is a nonlinear phase shift correction term determined by the nonlinear coefficient, pump power and polarization coupling coefficient.

[0045] In this embodiment, the quasi-phase matching condition does not require Δβ(z) to be equal to zero at a certain cross-section, but rather requires that the phase mismatch integral over the effective propagation length L of the silicon core fiber be within a preset range. Specifically, it can be made as follows: | _0^LΔβ(z)dz-2πm|≤ε Where m is the quasi-phase matching order and ε is the phase tolerance. For high selective frequency conversion, ε can be set to 0.1π to 0.2π; for more lenient experimental verification scenarios, ε can be set to 0.1π to 0.5π. The above values ​​are used to illustrate the selectable design range and can be adjusted according to the actual frequency conversion band, silicon core fiber loss, and pump conditions.

[0046] When establishing the joint constraint model, both loss and selectivity are considered. The effective action length Leff can be calculated based on the transmission loss. When the loss is approximately uniform, Leff = [1 - exp(-αL)] / α; when the loss varies with axial position, the equivalent action length can be obtained by integrating α(z) along the axial direction. The selectivity of the target idler light can be characterized by the side-mode suppression ratio (SMSR), where SMSR represents the ratio between the target idler light power and the strongest non-target sideband power. By incorporating conversion efficiency, side-mode suppression ratio, pump threshold, non-target sideband intensity, total loss, and processing deviation into the same evaluation function, the increased loss or enhanced non-target sidebands caused by solely pursuing a smaller silicon core diameter or zero dispersion point can be avoided.

[0047] After completing the joint constraint modeling, the axial diameter distribution d(z) of the tapered waist region 205 of the silicon core optical fiber was determined. See [link / reference needed]. Figure 3 The silicon-core fiber dispersion programming frequency conversion unit 110 includes an input-side silicon-core fiber segment 203, an input transition cone region 204, a tapered waist region 205, an output transition cone region 206, and an output-side silicon-core fiber segment 207. The input transition cone region 204 smoothly transitions the input-side silicon core diameter to the starting diameter of the tapered waist region 205, and the output transition cone region 206 smoothly transitions the end diameter of the tapered waist region 205 to the output-side silicon core diameter. The tapered waist region 205 can adopt a segmented stepped, continuously tapered, periodically modulated, or composite diameter distribution. For each candidate diameter distribution, its corresponding β(ω,z), γ(z), α(z), and Δβ(z) are calculated, and it is determined whether it meets the quasi-phase matching, loss, and selectivity requirements.

[0048] After determining the axial diameter distribution of the tapered waist region, the pump parameters and input polarization state are configured. In single-pump mode, the pump light wavelength λp is determined based on the target wavelength relationship, and the pump power range is set according to the nonlinear coefficients and loss parameters calculated by the model. In dual-pump mode, candidate combinations of λp1 and λp2 are determined based on the target wavelength relationship, and the frequency interval between the two pump lights is adjusted by the dual-pump spacing control module 107. Before the signal light and pump light enter the silicon core fiber, they can pass through the signal polarization controller 103 and the pump polarization controller 108 respectively, so that their effective nonlinear coupling state in the silicon core fiber meets the preset requirements.

[0049] Then, the signal light and pump light are input into the silicon core fiber dispersion-programmed frequency converter 110. See also Figure 2The signal light input unit 101 outputs the signal light to be converted, the pump light source unit 104 outputs single-pump or dual-pump pump light, and the beam combining unit 109 couples the signal light and pump light to the silicon core fiber dispersion-programmed frequency conversion unit 110. Within the tapered waist region 205, the signal light and pump light generate idler light through the third-order nonlinear interaction of the silicon core material. Because the tapered waist region 205 has a silicon core diameter distribution that varies along the light propagation direction, the phase mismatch accumulation corresponding to the target idler light can be controlled within a preset tolerance range, thereby allowing the target idler light to continuously accumulate within the effective propagation length.

[0050] Finally, the output spectrum is acquired and feedback correction is performed. The output filtering unit 112 separates the target idler light from the output light of the silicon core fiber, and the spectral monitoring unit 113 acquires the output spectrum. The feedback control unit 114 identifies the target idler spectral peak 702, the residual signal spectral peak 703, the residual pump spectral peak 704, and the non-target sideband 705 from the output spectrum, and calculates the conversion efficiency, side-mode suppression ratio, and the center wavelength offset of the target idler light. When the target idler light power is lower than a preset value, the feedback control unit 114 outputs a pump power adjustment command 709 or a polarization state adjustment command 711; when the center wavelength of the target idler light deviates from the preset value, the feedback control unit 114 outputs a temperature adjustment command 710 or a pump wavelength adjustment command; when the non-target sideband 705 exceeds a preset threshold, the feedback control unit 114 adjusts the fiber temperature, the dual-pump spacing, or the input polarization state to make the non-target sideband deviate from the phase matching condition. Through the above closed-loop control, this embodiment can compensate for the effects of tapered diameter processing deviation, temperature drift, and polarization state changes on the target idler light output. Example 2

[0051] like Figure 7 As shown, this embodiment illustrates the specific implementation of the joint constraint model 600. This model is used to determine the axial diameter distribution d(z), pump wavelength, dual-pump spacing, pump power, input polarization state, and temperature adjustment of the tapered waist region 205 of the silicon core fiber according to the target wavelength conversion task. This model can be used for structural design before fabrication of the silicon core fiber dispersion-programmable frequency converter unit 110, or for operating point search and feedback correction after device fabrication.

[0052] See Figure 7 First, the target wavelength input 601 receives the wavelength of the signal light to be converted, λs, and the target idler wavelength, λi. After receiving the target wavelength, the candidate pump combination generation unit 603 generates candidate pump wavelengths according to the energy conservation relationship. In single-pump mode, the candidate pump angular frequency ωp satisfies 2ωp=ωs+ωi. In dual-pump mode, the candidate pump angular frequencies ωp1 and ωp2 satisfy ωp1+ωp2=ωs+ωi. If the tunable range of the actual pump source is limited, the candidate pump combination also needs to meet the conditions of pump source wavelength range, output power range, and linewidth.

[0053] Subsequently, the parameter database 602 provides the model with silicon core fiber parameters under different structural and operational conditions. The parameter database 602 may include silicon core diameter d, cladding material refractive index, temperature T, effective refractive index neff, propagation constant β, group velocity dispersion β2, higher-order dispersions β3 and β4, effective mode area Aeff, nonlinear coefficient γ, and transmission loss α. To enable the model to handle structures with varying axial orientations, the parameter database 602 preferably stores parameters indexed by axial position z. Once the candidate diameter distribution d(z) is generated, the model can call the corresponding parameters at each axial sampling position, forming a complete sequence of axial parameters.

[0054] The candidate diameter distribution generation unit 604 generates multiple candidate d(z) based on the target wavelength conversion task. The candidate d(z) can be a piecewise stepped distribution, a continuously tapered distribution, a periodically modulated distribution, or a composite distribution. A piecewise stepped distribution can be represented by the diameter and length of multiple dispersion control segments; a continuously tapered distribution can be represented by a polynomial function, spline function, exponential function, or hyperbolic tangent function; and a periodically modulated distribution can be represented by the average diameter, modulation amplitude, and modulation period. The candidate diameter distribution generation unit 604 can also introduce fabrication feasibility constraints, such as the minimum silicon core diameter, the maximum diameter change rate, the minimum transition region length, and the maximum total insertion loss.

[0055] The propagation constant and dispersion calculation unit 605 calculates βp(z), βs(z), and βi(z) based on the candidate d(z). Here, βp(z) corresponds to the pump light propagation constant, βs(z) corresponds to the signal light propagation constant, and βi(z) corresponds to the target idler light propagation constant. When using a dual-pump mode, βp1(z) and βp2(z) are also calculated. The propagation constant can be calculated based on the effective refractive index, i.e., β(ω,z) = neff(ω,z)·ω / c, where c is the speed of light in vacuum. Group velocity dispersion and higher-order dispersion can be obtained from the second and higher-order derivatives of β with respect to the angular frequency, or directly from the dispersion parameters pre-stored in the parameter database 602.

[0056] In one specific implementation, neff(λ,d,T) is first obtained by solving the finite element model. Then, the propagation constant is calculated using β(ω,z) = neff(ω,z)·ω / c. β2 is obtained by taking the second derivative of β with respect to ω, and Aeff is obtained by integrating the mode field. The nonlinear coefficient is then obtained using γ = n2ω / (cAeff). Finally, α(z) is obtained using the cut-back method or the input-output power ratio. The step size for each axial sampling position is also specified, for example, from 0.1 mm to 1 mm.

[0057] The nonlinear coefficient and loss calculation unit 606 calculates γ(z) based on the effective mode area Aeff(z) and the nonlinear refractive index of the material. In one optional calculation method, γ(z) = n²·ωp / [c·Aeff(z)], where n² is the nonlinear refractive index of the silicon core material. Since Aeff(z) typically decreases as the silicon core diameter decreases, a smaller silicon core diameter is beneficial for increasing γ(z). However, when the silicon core diameter is too small, the transmission loss α(z) may increase, and additional losses may also occur in the transition cone region. Therefore, this embodiment calculates α(z) and the effective action length Leff obtained from α(z) simultaneously, avoiding limiting the candidate structure selection to simply the highest nonlinear coefficient.

[0058] The phase mismatch integral calculation unit 607 calculates Δβ(z) based on the propagation constant, nonlinear coefficient, and pump power, and further calculates... 0L Δβ(z)dz. For single-pump mode, Δβ(z)=2βp(z)-βs(z)-βi(z)+2γ(z)Pp(z). For dual-pump mode, Δβ(z)=βp1(z)+βp2(z)-βs(z)-βi(z)+ΔβNL(z). When the pump power decays along the axial direction due to losses, Pp(z), Pp1(z), and Pp2(z) can be calculated from the integral of the input power and axial losses. In an alternative approach, ΔβNL(z) can be calculated from γ(z)[κ1(z)Pp1(z)+κ2(z)Pp2(z)], where κ1(z) and κ2(z) are dimensionless coupling coefficients determined by the polarization relationship between the pump light, signal light, and idler light.

[0059] The objective function evaluation unit 608 is used to comprehensively evaluate candidate structures and candidate control parameters. The evaluation process can simultaneously consider conversion efficiency η, side-mode suppression ratio (SMSR), pump threshold Pth, non-target sideband intensity Rnoise, total loss (Loss_total), and processing deviation Derror. The objective function can be expressed as: F=w1·η+w2·SMSR-w3·Pth-w4·Rnoise-w5·Loss_total-w6·Derror Wherein, w1, w2, w3, w4, w5, and w6 are weighting coefficients. These weighting coefficients can be determined based on the application scenario. In communication wavelength conversion scenarios, the weights corresponding to SMSR and Loss_total can be increased; in extended band light source scenarios, the weights corresponding to η and Pth can be increased; and in scenarios with large fabrication tolerances, the weight corresponding to Derror can be increased. The above objective function is used to illustrate multi-objective constraint methods and does not limit this invention to using only this specific functional form.

[0060] When the candidate structure satisfies | When 0L Δβ(z)dz-2πm|≤ε, the total loss does not exceed the preset loss threshold, the target idler light conversion efficiency is not lower than the preset value, and the side-mode suppression ratio is not lower than the preset value, the optimal parameter output unit 609 outputs the corresponding axial diameter distribution of the tapered waist region, pump wavelength, dual-pump interval, pump power, polarization state, and temperature control quantity. When multiple candidate structures meet the above constraints, the scheme with lower total loss, smaller diameter change rate, larger processing tolerance, or lower pump power can be preferentially selected.

[0061] After the device fabrication is completed, this embodiment can also write back the actually measured d(z) to the parameter database 602. Specifically, the diameter distribution of the actual tapered waist region 205 is obtained through microscopic imaging, scanning electron microscopy, near-field mode testing, or output spectrum inversion, and then β(ω,z), γ(z), α(z), and Δβ(z) are recalculated. If the actual phase mismatch integral deviates from the design target, the feedback control unit 114 compensates for the deviation by adjusting the pump wavelength, dual-pump interval, temperature, or polarization state. In this way, the joint constraint model is not only used for the initial design but also for the correction of the operating point during the actual operation of the device. Example 3

[0062] This embodiment is for illustration. Figures 3 to 6 The diagram illustrates the structural implementation of the silicon core fiber dispersion-programmable frequency converter 110. The silicon core fiber dispersion-programmable frequency converter 110 includes an input transition cone region 204, a tapered waist region 205, and an output transition cone region 206. The tapered waist region 205 is the main nonlinear frequency conversion region, and its silicon core diameter distribution d(z) is determined according to the target wavelength conversion task. The input transition cone region 204 and the output transition cone region 206 are used to reduce mode change loss and enable the optical signal to enter and leave the tapered waist region 205 more stably.

[0063] See Figure 3The silicon core fiber 200 includes a silicon core 201 located in the central region and a cladding 202 covering the outside of the silicon core 201. An input-side silicon core fiber segment 203, an input transition taper region 204, a tapered waist region 205, an output transition taper region 206, and an output-side silicon core fiber segment 207 are sequentially arranged along the light propagation direction 210. The input-side silicon core fiber segment 203 and the output-side silicon core fiber segment 207 are used for connection to external fiber optic links or optical devices; the input transition taper region 204 gradually reduces the diameter of the input-side silicon core to the initial diameter of the tapered waist region 205 to reduce additional losses caused by abrupt changes in input mode; the output transition taper region 206 gradually transitions the end diameter of the tapered waist region 205 to the diameter of the output-side silicon core to improve output mode matching. The tapered waist region 205 has an axial diameter distribution d(z) that varies along the light propagation direction. This axial diameter distribution is used to adjust the propagation constant β(ω,z), group velocity dispersion β2(z), nonlinear coefficient γ(z), and transmission loss α(z) at different axial positions. A temperature-controlled contact area 208 is disposed on the outside of the tapered waist region 205 to transfer heat generated by the temperature control unit 111 to the tapered waist region 205 or its adjacent area. A protective encapsulation layer 209 is disposed on the outside of the silicon core optical fiber 200 to fix the tapered, narrow-diameter structure and reduce the impact of mechanical disturbances on the phase-matching state. Figure 3 This illustration only shows the axial and functional relationships between the various structures and does not specify the actual proportions of each part.

[0064] See Figure 4 In one embodiment, the tapered waist region 205 adopts a segmented stepped tapered waist region 300. The segmented stepped tapered waist region 300 includes a first dispersion control segment 301, a second dispersion control segment 302, a third dispersion control segment 303, and an nth dispersion control segment 304, with short gradient transition regions 305 provided between adjacent dispersion control segments. Each dispersion control segment has a corresponding silicon core diameter, axial length, group velocity dispersion, nonlinear coefficient, and transmission loss. The stepped diameter distribution curve 306 is used to represent the segmented variation of the silicon core diameter d with the axial position z.

[0065] In the segmented stepped structure, the dispersion control section near the input end can use a relatively large silicon core diameter to reduce input mode perturbation; the intermediate dispersion control section can use a smaller silicon core diameter to reduce the effective mode area and improve the nonlinear coefficient; the dispersion control section near the output end can moderately increase the silicon core diameter to compensate for the accumulated phase error of the preceding stage and reduce output loss. For example, the tapered waist region 205 can include five dispersion control sections with silicon core diameters of 1.20 μm, 0.90 μm, 0.80 μm, 0.95 μm, and 1.10 μm, respectively, and the length of each section can be selected within the range of 1 mm to 20 mm. This structure enables different axial sections to respectively undertake the functions of input mode transition, nonlinear enhancement, phase error compensation, and output spectrum purification.

[0066] The segmented stepped structure can be fabricated using CO2 laser scanning tapering, flame-heated tapering, localized tapering in an electric resistance furnace, or tapering with a microheater. During fabrication, the diameter and length of each dispersion control segment can be adjusted by controlling the heating power, traction speed, scanning speed, and heating zone length. To reduce mode disturbances between stepped segments, a short gradient transition zone 305 is preferably provided between adjacent dispersion control segments. The length of the short gradient transition zone 305 can be determined based on the difference between adjacent diameters, typically ranging from 0.1 mm to 2 mm. This length range is an optional design range and can be adjusted according to the actual fiber material and tapering equipment.

[0067] See Figure 5 In another embodiment, the tapered waist region 205 adopts a continuously tapered waist region 400. The continuously tapered waist region 400 includes an input tapered region 401, an intermediate high nonlinearity region 402, and an output tapered region 403. The continuous diameter distribution curve 404 indicates that the silicon core diameter d changes continuously along the axial position z. The continuously tapered structure can avoid the abrupt diameter changes in the segmented stepped structure, which is beneficial to reducing scattering loss and mode conversion loss.

[0068] A continuously varying diameter distribution can be described by a polynomial function, spline function, exponential function, or hyperbolic tangent function. For example, d(z) can be expressed as d(z) = Spline(z; d1, d2, ..., dN), where d1 to dN are the silicon core diameters at several axial control points. It can also be expressed as d(z) = d0 + a1z + a2z 2 Where d0, a1, and a2 are determined by the joint constraint model. In the continuous gradient structure, the input gradient region 401 can gradually reduce the diameter of the larger silicon core to the target diameter of the intermediate high nonlinear interaction region 402, which provides strong nonlinear interaction. The output gradient region 403 gradually restores the diameter of the silicon core to a larger diameter to reduce output loss.

[0069] In a continuously variable structure, the diameter change rate can be used as a processing constraint input to the joint constraint model. If the diameter changes too rapidly, mode coupling and additional losses may occur; if the diameter changes too slowly, the length of the highly nonlinear active region may be insufficient. Therefore, the continuous diameter distribution curve 404 can be determined jointly based on the phase mismatch integral and the effective active length. For cross-band frequency conversion tasks, the continuously variable structure allows Δβ(z) to change slowly along the axial direction, enabling the target idler light to maintain a near-phase-matched state over a longer range.

[0070] See Figure 6In another embodiment, the tapered waist region 205 adopts a periodically modulated tapered waist region 500. The periodically modulated tapered waist region 500 has an average diameter line 501 and a periodic diameter modulation curve 502. The silicon core diameter d(z) can be expressed as: d(z)=d0+Δd·sin(2πz / Λ+φ0) Where d0 is the average silicon core diameter, Δd is the modulation amplitude 504, Λ is the modulation period 503, and φ0 is the initial phase. Periodic diameter modulation enables the local propagation constant and group velocity dispersion to change periodically along the axial direction, thereby forming the periodic compensation effect corresponding to the phase mismatch compensation schematic curve 505.

[0071] In the periodic modulation structure, the modulation period Λ can be determined based on the equivalent phase mismatch of the target idler light. When the equivalent phase mismatch of the target idler light is close to 2πq / Λ, periodic diameter modulation can provide an equivalent quasi-phase-matching condition, where q is the modulation order. By selecting d0, Δd, and Λ, the target idler light can achieve coherent accumulation over multiple modulation periods, while simultaneously causing non-target sidebands to deviate from the quasi-phase-matching condition of that period. This structure is suitable for tasks requiring improved selectivity of the target idler light or suppression of wavelength conversion in adjacent non-target sidebands.

[0072] The average diameter d0 of the periodically modulated structure can be selected from 0.6 μm to 2.5 μm, the modulation amplitude Δd can be selected from 10 nm to 300 nm, and the modulation period Λ can be selected from 0.1 mm to 10 mm. These ranges are selectable, and the specific values ​​are determined by the target wavelength, pump wavelength, silicon core material dispersion, cladding structure, and loss constraints. The periodically modulated structure can be fabricated by periodically changing the heating power, periodically changing the traction speed, or by using localized heat treatment.

[0073] In another embodiment, the tapered waist region 205 employs a composite diameter distribution. This composite diameter distribution can combine segmented stepped regions, continuous gradient regions, periodic modulation regions, and flat high nonlinear regions. For example, an input gradient region, a first periodic modulation region, a flat high nonlinear region, a second periodic modulation region, and an output gradient region can be sequentially configured. The input gradient region is used to reduce input loss, the first periodic modulation region is used to establish quasi-phase matching conditions for the target idler light, the flat high nonlinear region is used to accumulate the target idler light power, the second periodic modulation region is used to compensate for phase mismatch changes after pump attenuation, and the output gradient region is used to reduce output mode perturbations. This composite structure is suitable for frequency conversion tasks with large wavelength spans, strict loss constraints, or the need to accommodate multiple target idler light channels.

[0074] To ensure that the actual structure of the tapered waist region 205 meets the model design requirements, structural and optical verification can be performed during or after fabrication. Structural verification can be performed by obtaining d(z) using optical microscopy, scanning electron microscopy, or online diameter monitoring. Optical verification can be performed by transmission loss testing, near-field mode testing, white light interferometry testing, output spectrum scanning, and temperature scanning. If the measured d(z) deviates from the design value, the measured d(z) can be re-entered into the parameter database 602, and the pump wavelength, pump power, fiber temperature, or input polarization state can be adjusted through the feedback control unit 114 to compensate for the phase matching error introduced by the processing deviation.

[0075] This embodiment illustrates various dispersion programming implementation methods for the tapered waist region 205 through segmented stepped, continuously varied, periodically modulated, and composite structures. The common feature of these structures is that they all control the variation of the silicon core diameter d with the axial position z to form a preset distribution of the propagation constant, group velocity dispersion, nonlinear coefficient, and transmission loss along the light propagation direction. This ensures that the accumulated phase mismatch corresponding to the target idler light meets the quasi-phase matching condition and suppresses non-target sidebands. Example 4

[0076] like Figure 2 As shown, this embodiment illustrates the specific composition and operation of the dispersion-programmed frequency conversion device 100 based on silicon core optical fiber. This device 100 executes the aforementioned dispersion-programmed frequency conversion method based on silicon core optical fiber, and achieves target idler frequency output through signal light input, pump light input, optical path combining, silicon core optical fiber nonlinear frequency conversion, temperature adjustment, spectral monitoring, and feedback control. The functional modules in the device 100 can be constructed using independent devices, or some modules can be integrated into the same optical platform, the same control platform, or the same packaging structure, as long as the modules can achieve corresponding optical path connections, control connections, and data interaction.

[0077] See Figure 2 The device 100 includes a signal light input unit 101, a pump light source unit 104, a beam combining unit 109, a silicon core fiber dispersion programming frequency conversion unit 110, a temperature control unit 111, an output filtering unit 112, a spectrum monitoring unit 113, and a feedback control unit 114. The device 100 may also include a signal polarization controller 103, a pump polarization controller 108, a dual-pump interval control module 107, a target idler light output terminal 115, and a residual light output or monitoring terminal 116. Figure 2 In the diagram, optical path connection 118 is used to indicate the transmission path of signal light, pump light, target idler light or residual light, and control signal connection 117 is used to indicate the control relationship between feedback control unit 114 and pump light source unit 104, temperature control unit 111, polarization controller and dual pump interval control module 107.

[0078] The signal light input unit 101 provides the signal light to be converted. This signal light can originate from a tunable laser, a communication transmitter, the output of an external modulator, the output of an optical fiber link, or a test light source. The signal light input unit 101 may include a signal light source 102, an optical isolator, an adjustable optical attenuator, an optical amplifier, a signal polarization controller 103, and a wavelength monitor. The signal light source 102 generates or receives signal light with a wavelength of λs; the adjustable optical attenuator adjusts the input signal light power to ensure it is within the range allowed by the silicon-core fiber dispersion programming frequency converter 110; the optical isolator reduces the impact of reflected light on the preceding light source; and the signal polarization controller 103 adjusts the polarization state of the signal light entering the silicon-core fiber. For signal light carrying modulation information, the signal light input unit 101 does not change its modulation format, but only adapts the input power, input polarization state, and input wavelength.

[0079] Pump source unit 104 provides pump light for nonlinear frequency conversion. Pump source unit 104 may include a first pump source 105 and may further include a second pump source 106. In single-pump mode, the first pump source 105 outputs pump light with wavelength λp, which is input together with the signal light into the silicon core fiber dispersion programming frequency conversion unit 110. In dual-pump mode, the first pump source 105 and the second pump source 106 output first and second pump lights with wavelengths λp1 and λp2 respectively, and their wavelength or frequency interval is adjusted by the dual-pump interval control module 107. The first pump source 105 and the second pump source 106 can be narrow-linewidth continuous-wave lasers, tunable semiconductor lasers, external cavity lasers, fiber lasers, or pulsed pump sources. For transparent frequency conversion of communication signals, narrow-linewidth continuous-wave pump sources or quasi-continuous-wave pump sources are preferred; for extended-band light generation, pulsed pump sources can be used to increase peak power.

[0080] The pump source unit 104 may further include a pump power regulator, a pump polarization controller 108, a pump optical isolator, and a pump wavelength monitor. The pump power regulator adjusts the pump optical power according to control commands output by the feedback control unit 114. The pump polarization controller 108 adjusts the polarization state of the pump light entering the silicon core fiber, ensuring an effective nonlinear coupling state suitable for the four-wave mixing process between the pump light and the signal light. The pump wavelength monitor monitors the actual values ​​of λp, λp1, or λp2 and feeds the monitoring results back to the feedback control unit 114. The dual-pump interval control module 107 can be implemented by adjusting the output wavelengths of the first pump source 105 and the second pump source 106 respectively, or by controlling the frequency interval between the two pump lights through modulation, filtering, or phase-locked loop.

[0081] The beam combiner 109 couples the signal light and pump light into the silicon-core fiber dispersion-programmed frequency converter 110. The beam combiner 109 can be a wavelength division multiplexer, a broadband fiber coupler, a polarization beam combiner, a free-space beam combiner, or an integrated optical beam combiner. When the signal light and pump light are in different wavelength bands, the beam combiner 109 preferably uses a wavelength division multiplexer; when the wavelengths of the signal light and pump light are similar, a broadband coupler or a polarization beam combiner can be used. The output of the beam combiner 109 is connected to the input of the silicon-core fiber dispersion-programmed frequency converter 110, allowing the signal light and pump light to enter the silicon-core fiber along the same optical path.

[0082] The silicon-core fiber dispersion-programmable frequency converter 110 is a nonlinear frequency conversion component in device 100 that generates target idler light. See also... Figure 3 The silicon-core fiber dispersion-programmable frequency conversion unit 110 includes a silicon-core fiber 200, which has a silicon core 201 and a cladding 202. Along the light propagation direction, it is sequentially configured with an input-side silicon-core fiber segment 203, an input transition cone region 204, a tapered waist region 205, an output transition cone region 206, and an output-side silicon-core fiber segment 207. The input transition cone region 204 gradually transitions the silicon core diameter of the input-side silicon-core fiber segment 203 to the initial diameter of the tapered waist region 205. The output transition cone region 206 gradually transitions the end diameter of the tapered waist region 205 to the silicon core diameter of the output-side silicon-core fiber segment 207. The tapered waist region 205 has a silicon core diameter distribution d(z) that varies along the light propagation direction, which is determined by the target wavelength conversion task and the dispersion-nonlinearity-loss joint constraint model.

[0083] In specific implementations, the silicon core fiber dispersion-programmable frequency conversion unit 110 can adopt a segmented stepped tapered waist region 300, a continuously tapered waist region 400, a periodically modulated tapered waist region 500, or a composite tapered waist region. The segmented stepped tapered waist region 300 includes multiple dispersion control segments, each corresponding to a set of propagation constants, group velocity dispersion, nonlinear coefficients, and transmission losses. The continuously tapered waist region 400 causes the silicon core diameter to change continuously along the axial direction to reduce mode perturbations. The periodically modulated tapered waist region 500 causes the silicon core diameter to change periodically around the average diameter line 501 according to the modulation period 503 and modulation amplitude 504 to provide equivalent quasi-phase matching conditions. The composite tapered waist region can combine at least two of the above structures to adapt to cross-band frequency conversion, low-loss frequency conversion, or multi-channel target idler optical output scenarios.

[0084] The silicon core fiber dispersion programming frequency converter unit 110 can be equipped with a protective encapsulation layer 209. The protective encapsulation layer 209 is used to fix the tapered silicon core fiber and reduce the impact of environmental vibration, airflow disturbance, and mechanical stress on the tapered waist region 205. The protective encapsulation layer 209 can be formed using a quartz tube, metal groove, ceramic substrate, or transparent encapsulation material. To facilitate temperature control, a temperature control contact area 208 can be provided on the outer side of the tapered waist region 205. The temperature control contact area 208 is in thermal contact with the temperature control unit 111, enabling the temperature control unit 111 to adjust the local or overall temperature of the silicon core fiber dispersion programming frequency converter unit 110.

[0085] The temperature control unit 111 is used to regulate the temperature of the silicon core fiber dispersion programming frequency converter unit 110. The temperature control unit 111 may include a thermoelectric cooler, a micro-heating element, a metal heat-conducting base, a temperature sensor, a heat insulation structure, and a temperature controller. The temperature control unit 111 can adopt a unified temperature control method, that is, apply a uniform temperature to the entire silicon core fiber dispersion programming frequency converter unit 110; or it can adopt a zoned temperature control method, that is, set multiple temperature control zones along the silicon core fiber axis. The zoned temperature control method can separately adjust the temperature of the main nonlinear action regions in the input transition cone region 204, the tapered waist region 205, and the output transition cone region 206, thereby more precisely compensating for local phase mismatch deviations. The temperature control unit 111 receives the temperature adjustment command 710 output by the feedback control unit 114 and adjusts the temperature according to the target idler wavelength offset or the non-target sideband intensity.

[0086] Output filtering unit 112 is used to separate the target idler light from the mixed light output from silicon core fiber dispersion programming frequency conversion unit 110. The output mixed light typically includes residual signal light, residual pump light, target idler light, and non-target sidebands. Output filtering unit 112 can be a tunable narrowband optical filter, wavelength division multiplexer, fiber Bragg grating, arrayed waveguide grating, thin film filter, or free space filter. After passing through output filtering unit 112, the target idler light is output from target idler light output terminal 115. The residual pump light, residual signal light, and some non-target sidebands can be output through residual light output or monitoring terminal 116, or guided to spectral monitoring unit 113 for monitoring.

[0087] The spectral monitoring unit 113 is used to acquire the output spectrum 701. The spectral monitoring unit 113 can employ a structure combining a spectral analyzer, a grating-type miniature spectrometer, an array detector, a tunable filter, and a photodetector, or a structure combining a wave demultiplexer and multiple photodetectors. The spectral monitoring unit 113 can at least identify the target idler spectral peak 702, the residual signal spectral peak 703, the residual pump spectral peak 704, and non-target sidebands 705. For devices requiring rapid feedback control, instead of acquiring the complete spectrum, multiple monitoring channels can be set at the corresponding wavelengths of the target idler light, the residual pump light, and the main non-target sidebands to reduce control delay.

[0088] The feedback control unit 114 generates control commands based on the spectral information output by the spectral monitoring unit 113. The feedback control unit 114 can be a microprocessor, a field-programmable gate array (FPGA), an industrial controller, a host computer, or an embedded control board. The feedback control unit 114 can store a silicon core fiber parameter database 602, candidate pump combinations, candidate diameter distributions, objective functions, historical operating points, and control parameters. The feedback control unit 114 calculates the conversion efficiency, side-mode suppression ratio, and target idler wavelength offset based on the output spectrum 701, and outputs pump power or pump wavelength adjustment commands to the pump light source unit 104, a temperature adjustment command 710 to the temperature control unit 111, a polarization state adjustment command 711 to the signal polarization controller 103 or pump polarization controller 108, or a dual-pump interval adjustment command to the dual-pump interval control module 107.

[0089] See Figure 8 The feedback control process can be executed as follows: After the spectral monitoring unit 113 acquires the output spectrum 701, the feedback control unit 114 identifies the position and power of the target idler spectral peak 702. If the target idler optical power is lower than a preset power threshold, the feedback control unit 114 can first adjust the pump power; if the target idler optical power is still insufficient after adjusting the pump power, the input polarization state is further adjusted to improve the effective nonlinear coupling state. If the center wavelength of the target idler optical deviates from the target wavelength, the feedback control unit 114 can adjust the pump wavelength, the dual-pump interval, or the set temperature of the temperature control unit 111. If the power of the non-target sideband 705 exceeds a preset threshold, the feedback control unit 114 can adjust the fiber temperature, the dual-pump interval, or the input polarization state to make the non-target sideband deviate from the phase matching condition. If the conversion efficiency, side-mode suppression ratio, and center wavelength of the target idler optical all meet the preset requirements, the feedback control unit 114 locks the current pump power, pump wavelength, temperature, and polarization state, so that the device 100 is in a stable output state.

[0090] In this embodiment, each module of the device 100 corresponds to a method flow. The signal light input unit 101 corresponds to the step of acquiring and inputting the signal light to be converted; the pump light source unit 104 and the dual-pump interval control module 107 correspond to the step of determining the pump wavelength and the dual-pump interval; the feedback control unit 114 and its internal parameter database correspond to the step of constructing a joint constraint model and outputting control parameters; the silicon core fiber dispersion programming frequency conversion unit 110 corresponds to the step of realizing nonlinear frequency conversion according to the axial diameter distribution; the spectrum monitoring unit 113 and the output filtering unit 112 correspond to the steps of output spectrum acquisition and target idler light extraction; the temperature control unit 111, the polarization controller, and the pump adjustment unit correspond to the feedback correction step. Through the above module correspondence, the device 100 can execute the dispersion programming frequency conversion method of the present invention. Example 5

[0091] like Figure 9 As shown, this embodiment illustrates an application scenario and output spectrum for extended band optical communication. This embodiment uses the conversion of C-band signal light to L-band target idler light as an example. This application scenario demonstrates the specific implementation process of this invention in optical fiber communication band relocation, including technical requirements, hardware selection, parameter settings, control optimization processing flow, and output effects. The parameters in this embodiment are optional design parameters used to illustrate the implementation of this invention and do not constitute a limitation on this invention.

[0092] To facilitate the explanation of the impact of the axial diameter distribution of the tapered waist region and the output spectrum feedback control on the target idler light output in this invention, this embodiment also provides numerical simulation comparison results based on a silicon core fiber parameter database and a dispersion-nonlinearity-loss joint constraint model. These numerical simulation comparison results are used to illustrate the parameter variation trends under different tapered structures and different feedback control methods, and do not imply that the same values ​​must be achieved in all embodiments of this invention. The conversion efficiency, side-mode rejection ratio, target idler light center wavelength deviation, and output power fluctuations of the actual device will be affected by factors such as the quality of the silicon core material, the processing accuracy of the tapered diameter, the cladding material, the stability of the pump source, the accuracy of temperature control, the accuracy of polarization control, and the bandwidth of the output filter.

[0093] See Figure 9 The input communication signal 801 is a C-band signal light with a center wavelength set to λs = 1535 nm. The target output is an L-band target idler light 805 with a center wavelength set to λi = 1605 nm. In single-pump frequency conversion mode, the wavelength of the pump light input 802 is determined according to 2ωp = ωs + ωi. Substituting λs = 1535 nm and λi = 1605 nm into the energy conservation relationship, the pump light wavelength λp is approximately around 1569.6 nm. Therefore, in this embodiment, the first pump light source 105 is selected as a narrow-linewidth tunable continuous-wave pump light source covering 1568 nm to 1572 nm. The pump light power can be adjusted within the range of 50 mW to 800 mW, and the specific operating point is determined by the output spectrum feedback.

[0094] The silicon core fiber dispersion programming frequency converter module 804 in this embodiment includes a silicon core fiber dispersion programming frequency converter unit 110, a temperature control unit 111, an output filtering unit 112, a spectrum monitoring unit 113, and a feedback control unit 114. The silicon core fiber 200 is composed of a silicon core 201 and a glass cladding 202. The diameter of the silicon core before tapering can be 5 μm to 10 μm, and the tapered waist region 205 adopts a segmented stepped diameter distribution. Five dispersion control segments can be sequentially set in the tapered waist region 205, with diameters of approximately 1.20 μm, 0.90 μm, 0.80 μm, 0.95 μm, and 1.10 μm, and lengths of each segment can be set to 5 mm, 6 mm, 8 mm, 6 mm, and 5 mm, respectively. The lengths of the input transition tapered region 204 and the output transition tapered region 206 can be set to 8 mm to 12 mm, respectively, to reduce input-output mode abrupt loss.

[0095] The design logic of the segmented stepped diameter distribution is as follows: the first dispersion control segment 301 uses a larger diameter to ensure that the signal light and pump light enter the frequency conversion region more smoothly; the second dispersion control segment 302 and the third dispersion control segment 303 use smaller diameters to reduce the effective mode field area and increase the nonlinear coefficient; the fourth and fifth dispersion control segments use moderately increased diameters to adjust the phase mismatch in the later stages and reduce the mode disturbance at the output end. The propagation constant, group velocity dispersion, nonlinear coefficient, and loss corresponding to each segment are calculated using the joint constraint model 600, and the phase mismatch integral corresponding to the target idler light is made to satisfy | 0L Δβ(z)dz-2πm|≤ε.

[0096] In terms of hardware connectivity, the C-band signal light, generated by the signal source 102, passes through the signal polarization controller 103 and the adjustable optical attenuator before entering the beam combiner unit 109. The pump light, generated by the first pump source 105, passes through the pump power regulator and the pump polarization controller 108 before entering the beam combiner unit 109. The beam combiner unit 109 couples the signal light and pump light and then inputs them into the silicon core fiber dispersion programming frequency conversion unit 110. The output filtering unit 112 selects the target idler light near 1605 nm and outputs it to the target idler light output terminal 115. The spectral monitoring unit 113 monitors the position and power of the residual signal light, residual pump light, target idler light, and non-target sidebands in the output spectral schematic region 806.

[0097] Regarding parameter settings, the signal light input power can be set within the range of -20 dBm to 20 dBm, with the specific value determined based on the communication signal format, system receiving sensitivity, and nonlinear frequency conversion efficiency. The initial wavelength of the pump source is set to around 1569.6 nm, and the initial pump power is set to a low value and gradually increased to avoid excessive pump power causing non-target sideband enhancement or temperature drift. The initial temperature of the temperature control unit 111 can be set to 25°C, with an adjustable range of 20°C to 80°C. The signal polarization controller 103 and the pump polarization controller 108 first perform a low-speed scan to find a polarization combination with high target idler light power, and then the feedback control unit 114 locks this polarization state.

[0098] During the control optimization process, the feedback control unit 114 first calls up the historical operating point or model calculation results corresponding to λs = 1535 nm and λi = 1605 nm, and sets the pump light wavelength, pump power, temperature, and polarization state to initial values. Subsequently, the spectral monitoring unit 113 acquires the output spectrum 701. If the target idler spectral peak 702 near 1605 nm does not appear in the spectrum, the feedback control unit 114 fine-tunes the pump light wavelength in the range of 1568 nm to 1572 nm, while slowly adjusting the set temperature of the temperature control unit 111 in the range of 20°C to 80°C until the target idler spectral peak 702 appears. If the target idler spectral peak 702 appears but the power is lower than the preset value, the feedback control unit 114 gradually increases the pump power and synchronously adjusts the input polarization state to improve effective nonlinear coupling.

[0099] Once the target idler spectrum peak 702 reaches a preset power, the feedback control unit 114 further calculates the side-mode suppression ratio. The side-mode suppression ratio can be determined based on the power difference between the target idler spectrum peak 702 and the strongest non-target sideband 705. If the side-mode suppression ratio is lower than the preset value, the feedback control unit 114 can reduce the pump power, fine-tune the fiber temperature, or fine-tune the pump light wavelength to make the non-target sideband 705 deviate from the phase-matching condition. If the center wavelength of the target idler light shifts relative to 1605 nm, the feedback control unit 114 can prioritize adjusting the pump wavelength; if the pump wavelength adjustment range is insufficient, the temperature is further adjusted to compensate for the propagation constant deviation using the thermo-optical effect.

[0100] During the stable output phase, the feedback control unit 114 maintains the pump power, pump wavelength, temperature, and polarization state near the locked operating point. When the spectral monitoring unit 113 detects a slow decrease in the target idler light power, the feedback control unit 114 first determines whether the residual pump spectral peak 704 has changed; if the residual pump light power decreases, the pump power is adjusted; if the residual pump light is stable while the target idler light decreases, polarization state fine-tuning is performed; if the center wavelength of the target idler light drifts, temperature fine-tuning is performed. In this way, the device 100 can maintain the target idler light output even in the presence of ambient temperature fluctuations, mechanical disturbances, or pump power fluctuations.

[0101] In this embodiment, the output spectrum schematic region 806 can display four types of spectral characteristics. The residual signal spectral peak 703 is located near 1535 nm, the residual pump spectral peak 704 is located near 1569.6 nm, the target idler spectral peak 702 is located near 1605 nm, and the non-target sideband 705 is located at other wavelengths. The target idler light 805 near 1605 nm can be extracted through the output filtering unit 112. If the system is set to a communication application, the allowable deviation of the center wavelength of the target idler light can be set to no more than ±0.3 nm, the side-mode rejection ratio can be set to no less than 15 dB or 20 dB, and the conversion efficiency threshold can be set according to the actual system power budget. The above indicators are design targets and can be adjusted according to device fabrication quality, pump power, signal modulation format, and system application requirements.

[0102] In another optional application scenario, the input communication signal 801 can be set to O-band signal light, for example, λs = 1310 nm, and the target idler light 805 can be set to C-band signal, for example, λi = 1550 nm. In this case, a dual-pump mode can be used, with the first pump light source 105 and the second pump light source 106 jointly providing the pump light required for frequency conversion. The dual-pump interval control module 107 determines and adjusts the frequency interval between the two pump lights according to ωp1 + ωp2 = ωs + ωi. The tapered waist region 205 can adopt a continuously tapered diameter distribution to reduce cross-band mode mismatch and additional losses. The temperature control unit 111 can adopt a zoned temperature control method to adjust the temperature of the input region, the intermediate nonlinear action region, and the output region separately to compensate for propagation constant deviations under different bands.

[0103] In another alternative application scenario, the input communication signal 801 can be set as a C-band signal light, and the target idler light 805 can be set as an extended band light near 2 μm. In this case, the pump light input 802 can select a pump source located in the range of 1.7 μm to 1.9 μm, and the tapered waist region 205 can adopt a periodically modulated diameter distribution. The periodically modulated structure provides an equivalent quasi-phase matching condition through the modulation period 503 and modulation amplitude 504, thereby enhancing the target idler light near 2 μm. The output filtering unit 112 can use a filter adapted to the 2 μm band, and the spectral monitoring unit 113 can use an extended band spectrometer or a detector of the corresponding band. This application scenario can be used for extended band light source generation, shortwave infrared detection, or related spectral measurements. When the target idler light is above 2.2 μm, the cladding material is preferably low-hydroxyl quartz, fluoride glass, chalcogenide glass, or other materials with low transmission loss in the target band.

[0104] In this embodiment, to illustrate the impact of the dispersion programming structure and output spectral feedback control on the target idler light output, a control scheme can be set up for simulation verification under the same input wavelength and pump conditions. During numerical simulation, the wavelength of the signal light to be converted, λs, is set to 1535 nm, the wavelength of the target idler light, λi, is set to 1605 nm, the wavelength of the single-pump pump light, λp, is set to 1569.6 nm, the input power of the signal light is set to 0 dBm, the input power of the pump light is set to 300 mW, the total length of the tapered waist region of the silicon core fiber is set to 30 mm, and the initial ambient temperature is set to 25℃. In the simulation, the signal light, pump light, and idler light are all considered to propagate along the fundamental or quasi-fundamental mode of the silicon core fiber, and the phase mismatch and the output power of the target idler light are calculated based on the effective refractive index, propagation constant, group velocity dispersion, effective mode area, nonlinear coefficient, and transmission loss at each axial sampling position. The axial sampling step size can be set from 0.1 mm to 0.5 mm. At each sampling position, β(ω,z), γ(z), and α(z) are obtained by calling or interpolating based on the silicon core diameter d(z) corresponding to that position.

[0105] The silicon core material used in the simulation can be low-defect silicon, and the cladding material can be low-hydroxyl quartz glass or borosilicate glass. For simulations from C-band to L-band, the refractive index of the silicon core can be obtained from the refractive index dispersion relationship of silicon material in the communication band, and the refractive index of the cladding can be obtained from the refractive index dispersion relationship of the corresponding glass material; the nonlinear refractive index n² of the silicon core material can be taken as 4.0 × 10⁻¹⁸ m. 2 / W up to 6.0×10^-18 m 2 The value within the range of / W is preferably 4.5 × 10^-18 m. 2 / W is used as a calculation example. The thermo-optical coefficient of silicon material can be taken in the range of 1.6×10^-4 / K to 2.0×10^-4 / K, and the thermo-optical coefficient of cladding material can be taken in the range of 0.8×10^-5 / K to 1.2×10^-5 / K. The above material parameters are used to establish the refractive index-temperature relationship under simulation conditions and do not constitute a limitation on the material parameters of this invention.

[0106] Regarding loss settings, the simulation represents the axial transmission loss of the silicon core fiber as α(z,λ), and varies it with the silicon core diameter and wavelength. For sections with a silicon core diameter of not less than 0.90 μm, the transmission loss can be set from 0.8 dB / cm to 1.5 dB / cm; for highly nonlinear sections with a silicon core diameter of approximately 0.80 μm, the transmission loss can be set from 1.2 dB / cm to 2.5 dB / cm; additional losses of 0.2 dB to 0.8 dB can be set for the input and output transition cone regions, respectively. The axial attenuation of the pump power can be expressed as Pp(z) = Pp(0)·exp[- The calculation is performed using αp(u)du], where αp(u) is the loss coefficient of the pump light at axial position u. The axial attenuation of the signal light and idler light can also be handled in the same way.

[0107] Regarding the temperature model, the simulation uses 25℃ as the initial temperature T0 and calculates the refractive index change caused by temperature change according to n(λ,T)=n(λ,T0)+(dn / dT)(T-T0). In the control scheme without feedback control, the silicon core fiber temperature is maintained at 25℃; in the scheme of this application with feedback control enabled, the feedback control unit adjusts the set temperature of the temperature control unit according to the deviation of the target idler light center wavelength. For the overall temperature control method, the same temperature correction amount is used at each axial sampling position in the tapered waist region; for the zoned temperature control method, different temperature correction amounts can be used for different dispersion control segments. The feedback simulation in this embodiment uses the overall temperature control method to illustrate the correction effect of feedback control on the target idler light center wavelength.

[0108] Regarding pump stability settings, the pump source is configured as a continuous-wave narrow-linewidth pump source, with an initial pump wavelength stability of no more than ±0.01 nm and short-term pump power fluctuation of no more than ±0.5%. The short-term signal light input power fluctuation is no more than ±0.2 dB. In the initial simulation, the polarization state is set to a state that ensures high effective nonlinear coupling between the signal light and the pump light. In the feedback control simulation, the polarization state is used as an adjustable control variable to further improve the target idler light power and side-mode suppression ratio after the target idler light appears.

[0109] Regarding evaluation metrics, the target idler light center wavelength deviation is defined as the difference between the center wavelength of the simulated target idler light spectrum peak and 1605 nm; the normalized conversion efficiency ηdB is defined as ηdB=10log10(Pi,out / Ps,in), where Pi,out is the target idler light output power and Ps,in is the signal light input power; the side-mode suppression ratio (SMSR) is defined as SMSR=10log10(Pi,target / Pi,side), where Pi,target is the target idler light spectrum peak power and Pi,side is the strongest non-target sideband power; the output power fluctuation is the maximum deviation of the target idler light output power relative to the average value within a set time window or a set disturbance range. The phase mismatch integral deviation is expressed in units of π and is used to reflect | 0LΔβ(z)dz-2πm| normalized deviation relative to π.

[0110] To ensure comparability of the comparison results, all three schemes used the same signal light wavelength, pump light wavelength, signal light input power, pump light input power, total length of the tapered waist region, and initial temperature. The only differences between the three schemes were whether the axial diameter distribution of the tapered waist region was dispersion-programmed and whether output spectral feedback control was enabled. Scheme 1 used a uniform tapered waist region with a fixed silicon core diameter of 0.90 μm; Scheme 2 used the same five-segment stepped tapered waist region as this application, but did not enable output spectral feedback correction; the scheme in this application used a five-segment stepped tapered waist region and corrected the pump power, fiber temperature, and input polarization state based on output spectral feedback. The simulation comparison results are shown in Table 1.

[0111] Table 1. Simulation comparison data based on parameter database and joint constraint model

[0112]

[0113] The comparison results above show that, under the same signal light, pump light, and effective operating length, the phase matching condition of the uniform tapered waist region, due to its fixed diameter, is mainly determined by the single-section dispersion. This results in a large phase mismatch integral deviation for the target idler light, leading to low conversion efficiency and easy enhancement of non-target sidebands. By adopting a segmented stepped tapered waist region, the propagation constant, group velocity dispersion, and nonlinear coefficients of each dispersion control segment change along the axial direction, compensating for the accumulated phase mismatch and improving the conversion efficiency and side-mode suppression ratio of the target idler light. Furthermore, by employing output spectrum feedback control, the feedback control unit corrects the pump power, fiber temperature, and input polarization state based on the target idler light center wavelength deviation, target idler light power, and non-target sideband power, making the target idler light center wavelength closer to the preset value and reducing output power fluctuations.

[0114] In a simulation feedback process, the initial output spectrum shows a target idler light center wavelength of 1604.58 nm and a sidemode suppression ratio (SMR) of 16.3 dB relative to the strongest non-target sideband. The feedback control unit first adjusts the set temperature of the temperature control unit from 25°C to 42°C, shifting the target idler light center wavelength to 1604.91 nm. Then, the pump light wavelength is fine-tuned from 1569.60 nm to 1569.68 nm, shifting the target idler light center wavelength to 1605.03 nm. Finally, the pump polarization controller fine-tunes the input polarization state of the pump light, increasing the target idler light power and raising the SMR to 23.1 dB. This process demonstrates that output spectrum feedback can compensate for phase matching deviations caused by tapered machining errors, temperature drift, and polarization state changes.

[0115] With output spectrum feedback control enabled, the feedback control unit sequentially corrects the temperature, pump wavelength, and polarization state based on the judgment results of the target idler light center wavelength deviation, conversion efficiency, and side-mode suppression ratio. Simulation data during the feedback process are shown in Table 2.

[0116] Table 2 Simulation data of the output spectrum feedback control process

[0117]

[0118] The data in the table above are used to illustrate the parameter variation trends of the technical solution of this application under simulation conditions, and do not limit the present invention to achieving the same values. In practical applications, the target idler light conversion efficiency, side-mode suppression ratio, and center wavelength deviation are affected by factors such as silicon core material quality, tapered diameter control accuracy, cladding structure, pump beam width, pump power stability, output filter bandwidth, and spectral monitoring accuracy. As long as the phase mismatch accumulation is adjusted by controlling the axial diameter distribution of the tapered waist region, and the pump power, fiber temperature, and input polarization state are corrected by output spectral feedback to improve the selectivity and stability of the target idler light output, this falls under the technical implementation described in this embodiment.

[0119] The data in Tables 1 and 2 are numerical simulation results obtained based on the aforementioned parameter database, loss settings, temperature model, and pump stability conditions. They are used to illustrate the trends in target idler light center wavelength deviation, conversion efficiency, side-mode rejection ratio, and output power fluctuation of the proposed scheme compared to the uniform tapered waist region scheme and the feedback-free control scheme. The above data does not constitute a limitation on the scope of protection of this invention, nor does it imply that the invention must achieve exactly the same values ​​under different material systems, different tapering processes, different pump powers, or different target wavelengths. For the actually fabricated silicon core fiber dispersion-programmable frequency converter, the parameter database can be regenerated based on the measured d(z), measured loss α(z), actual pump power stability, and actual temperature control accuracy, and the operating point can be re-determined through the output spectrum feedback process.

[0120] As can be seen from the above application scenarios, the method and apparatus of this invention are not limited to a fixed signal wavelength, a fixed pump wavelength, or a certain type of tapered structure. As long as the wavelength of the signal light to be converted and the target idler wavelength are used as inputs, and the axial diameter distribution and control parameters of the tapered waist region are determined through a silicon core fiber parameter database and a dispersion-nonlinearity-loss joint constraint model, and the pump power, temperature, and polarization state are corrected through output spectral feedback, dispersion-programmed frequency conversion of silicon core fibers for different band conversion tasks can be achieved.

[0121] In actual device operation, the output effect of the target idler light is not only determined by the nonlinearity of the silicon core material, but also by the tapered diameter distribution, cladding structure, pump wavelength, pump power, temperature, polarization state, and output filtering conditions. By incorporating the above factors into joint constraints and feedback control, this invention can systematically adjust the generation conditions of the target idler light and reduce output instability caused by processing errors, temperature drift, and polarization state changes.

[0122] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A dispersion-programmed frequency conversion method based on silicon core optical fiber, characterized in that, include: Acquire the wavelength of the signal light to be converted and the wavelength of the target idler light; The dispersion, nonlinear, and loss parameters of silicon-core optical fiber under different tapered diameters, cladding structures, pump wavelengths, and temperatures are obtained. A joint constraint model of dispersion, nonlinearity, and loss is constructed based on the wavelength of the signal light to be converted and the wavelength of the target idler light. The axial diameter distribution, pump wavelength, double-pump spacing, and input polarization state of the tapered waist region of the silicon-core optical fiber are determined based on the joint constraint model. The signal light and pump light are input into the silicon-core optical fiber to ensure that the signal light, pump light, and idler light satisfy the quasi-phase matching condition and generate the target idler light. The output spectrum is acquired, and at least one of the pump power, fiber temperature, and input polarization state is corrected based on the output spectrum.

2. The dispersion-programmed frequency conversion method based on silicon core optical fiber according to claim 1, characterized in that, The dispersion parameters include at least one of effective refractive index, propagation constant, group velocity dispersion, and higher-order dispersion; the nonlinear parameters include at least one of effective mode area and nonlinear coefficient; and the loss parameters include at least one of transmission loss, insertion loss, and additional loss introduced by the tapered transition region. When acquiring the dispersion, nonlinear, and loss parameters, the silicon core fiber is divided into multiple axial sampling positions along the light propagation direction. At each axial sampling position, the corresponding silicon core diameter, cladding equivalent refractive index, temperature, effective refractive index, nonlinear coefficient, and transmission loss are recorded or calculated, thereby forming a silicon core fiber parameter database.

3. The dispersion-programmed frequency conversion method based on silicon core optical fiber according to claim 1, characterized in that, The dispersion-nonlinearity-loss joint constraint model includes energy conservation constraints, axial phase mismatch constraints, loss constraints, and target idler light selectivity constraints. In single-pump frequency conversion mode, the pump light, signal light, and idler light satisfy 2ωp=ωs+ωi, and the axial phase mismatch is calculated according to Δβ(z)=2βp(z)-βs(z)-βi(z)+2γ(z)Pp(z). In dual-pump frequency conversion mode, the first pump light, second pump light, signal light, and idler light satisfy... Given ωp1+ωp2=ωs+ωi, and calculate the axial phase mismatch according to Δβ(z)=βp1(z)+βp2(z)-βs(z)-βi(z)+ΔβNL(z), where ΔβNL(z) is a nonlinear phase shift correction term determined by the nonlinear coefficient, the first pump power, the second pump power, and the polarization coupling coefficient, z is the axial position of the silicon core fiber, β is the propagation constant, γ is the nonlinear coefficient, and Pp, Pp1, and Pp2 are the pump powers.

4. The dispersion-programmed frequency conversion method based on silicon core optical fiber according to claim 3, characterized in that, The quasi-phase matching condition is that the axial phase mismatch integral within the effective propagation length of the silicon core fiber satisfies that the absolute deviation of the phase mismatch integral is no greater than the phase tolerance ε. _0^L Δβ(z)dz-2πm|≤ε, where L is the effective propagation length, m is the quasi-phase matching order, and ε is the phase tolerance; when determining the axial diameter distribution of the tapered waist region, the target idler light conversion efficiency, side-mode suppression ratio, pump threshold, non-target sideband intensity, total loss, and diameter processing deviation are simultaneously constrained to ensure that the phase mismatch integral corresponding to the target idler light is within the preset tolerance range, and that the non-target idler light deviates from the quasi-phase matching condition.

5. The dispersion-programmed frequency conversion method based on silicon core optical fiber according to claim 1, characterized in that, The axial diameter distribution of the tapered waist region of the silicon core optical fiber is segmented step, continuously tapered, periodically modulated, or a combination thereof; the segmented step axial diameter distribution includes at least two dispersion control segments with different silicon core diameters; the continuously tapered axial diameter distribution causes the silicon core diameter to continuously increase or decrease along the light propagation direction; the periodically modulated axial diameter distribution causes the silicon core diameter to periodically change around the average diameter along the light propagation direction; the combination axial diameter distribution includes at least two of the following: segmented step region, continuously tapered region, periodically modulated region, and flat high nonlinear region.

6. The dispersion-programmed frequency conversion method based on silicon core optical fiber according to claim 1, characterized in that, When correcting pump power, fiber temperature, and input polarization state based on output spectrum feedback, the target idler center wavelength, target idler power, residual pump power, residual signal power, and non-target sideband power are first identified from the output spectrum. Then, the conversion efficiency, sidemode rejection ratio, and target idler wavelength offset are calculated. When the target idler power is lower than a preset value, the pump power or input polarization state is adjusted. When the target idler center wavelength deviates from the preset wavelength, the pump wavelength, dual-pump spacing, or fiber temperature is adjusted. When the non-target sideband power exceeds a preset threshold, the fiber temperature, dual-pump spacing, or input polarization state is adjusted to make the non-target sideband deviate from the phase matching condition.

7. A dispersion-programmable frequency converter based on silicon core optical fiber, characterized in that, The system includes a signal light input unit, a pump light source unit, a polarization control unit, a beam combiner unit, a silicon core fiber dispersion-programmable frequency converter unit, a temperature control unit, an output filtering unit, a spectral monitoring unit, and a feedback control unit. The signal light input unit is used to input the signal light to be converted. The pump light source unit is used to output single-pump or dual-pump pump light. The polarization control unit is used to adjust the input polarization state of at least one of the signal light and the pump light. The beam combiner unit is used to couple the signal light and the pump light to the silicon core fiber dispersion-programmable frequency converter unit. The silicon core fiber dispersion programming frequency converter unit has a tapered waist region, and the tapered waist region has a silicon core diameter distribution that varies along the light propagation direction; the temperature control unit is used to adjust the temperature of the silicon core fiber dispersion programming frequency converter unit. The output filtering unit is used to separate the target idler light; the spectral monitoring unit is used to acquire the output spectrum. The feedback control unit is used to adjust at least one of the pump light source unit, temperature control unit, and input polarization state according to the output spectrum.

8. The dispersion-programmable frequency converter based on silicon core optical fiber according to claim 7, characterized in that, The silicon core fiber dispersion programming frequency conversion unit includes an input transition cone region, a tapered waist region, and an output transition cone region. The input transition cone region is used to transition the diameter of the input silicon core to the starting diameter of the tapered waist region, and the output transition cone region is used to transition the end diameter of the tapered waist region to the output silicon core diameter. The tapered waist region includes multiple dispersion control segments, or includes a continuously varying diameter distribution, or includes a periodically varying diameter distribution. Each dispersion control segment or each axial sampling position corresponds to a set of effective refractive index, group velocity dispersion, nonlinear coefficient, and transmission loss parameters, so that the tapered waist region forms a dispersion distribution in the light propagation direction to adjust the cumulative phase mismatch.

9. The dispersion-programmable frequency converter based on silicon core optical fiber according to claim 7, characterized in that, The feedback control unit stores a silicon core fiber parameter database and control parameters corresponding to the target wavelength conversion task. The silicon core fiber parameter database includes dispersion parameters, nonlinear parameters, and loss parameters under different tapered diameters, cladding structures, pump wavelengths, and temperature conditions. The feedback control unit calls or solves the axial diameter distribution of the tapered waist region, pump wavelength, dual-pump spacing, pump power, input polarization state, and temperature control quantities according to the wavelength of the signal light to be converted and the wavelength of the target idler light. The control unit corrects the control parameters when the target idler light power, center wavelength, or side-mode suppression ratio does not meet the preset conditions, based on feedback from the spectral monitoring unit.

10. The dispersion-programmable frequency converter based on silicon core optical fiber according to claim 7, characterized in that, The pump light source unit includes a first pump light source and a second pump light source, and a dual-pump interval control module is provided between the first pump light source and the second pump light source; the temperature control unit is an overall temperature control unit or a zoned temperature control unit; the polarization control unit is provided on at least one of the signal light input path and the pump light input path, and is used to adjust the effective nonlinear coupling state of the signal light and the pump light in the silicon core optical fiber.

Images