Optical waveguide adiabatic taper variable structure design method based on mode coupling loss slope control

CN122984186APending Publication Date: 2026-10-27NANTONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610364905.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-24
Publication Date
2026-10-27

AI Technical Summary

Technical Problem

[0009]本发明旨在解决现有光波导绝热锥变结构设计方法中存在的计算量大、数值稳定性差的问题,提供了基于模式耦合损耗斜率控制的光波导绝热锥变结构设计方法

Benefits of technology

[0026](1)计算效率极高:本方法的核心是计算损耗曲线的初始斜率,这是一个非迭代、非搜索的计算过程,其计算复杂度远低于寻找损耗峰值的过程。在相同的硬件条件下,本发明的设计速度可以比“峰值损耗法”提升一个数量级以上,极大地缩短了器件的研发周期。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122984186A_ABST
    Figure CN122984186A_ABST
Patent Text Reader

Abstract

The application belongs to the field of integrated optoelectronic technology, and particularly relates to a design method of an adiabatic taper variable structure of an optical waveguide based on mode coupling loss slope control. The application comprises the following steps: step one: initialization: defining initial geometric parameters, terminal geometric parameters, material refractive index and working wavelength λ of the taper variable waveguide; setting a target constant slope value S target ; setting an initial propagation step Δz; initializing the current position z as 0 and the current waveguide width W(z) as W in ; step two: local loss slope calculation; step three: local taper angle adjustment; step four: iteration; and step five: generating a final structure.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of integrated optoelectronics technology, specifically relating to a design method for an adiabatic tapered structure of an optical waveguide based on mode coupling loss slope control. Background Technology

[0002] In integrated optical circuits, the taper structure of an optical waveguide is a key component connecting waveguides of different widths, and is widely used in core devices such as mode size converters, spot size converters, couplers, and de)multiplexers. To achieve efficient optical power transmission, the taper structure must meet the adiabatic transmission condition, that is, during the continuous change of waveguide size, the fundamental mode energy hardly couples with higher-order modes, thereby avoiding mode conversion loss.

[0003] Traditionally, the design of adiabatic tapered structures has relied primarily on empirical profile functions, such as linear, parabolic, or exponential functions. While these functions are simple, they often fail to achieve the lowest possible loss within the shortest possible device length, resulting in wasted chip area.

[0004] To address this issue, the industry has proposed a more advanced "constant mode power loss" design criterion. Its typical implementation (hereinafter referred to as the "peak loss method" or TMLA) involves: at each point along the light propagation direction, calculating the mode coupling loss caused by local structural changes through numerical simulation, and finding the peak value of its loss oscillation curve. Then, through a feedback algorithm, the steepness of the taper at that point (i.e., the local taper angle) is dynamically adjusted so that this loss peak value remains a preset, very small, constant value over the entire taper length. This method can design taper structures with performance approaching the theoretical limit.

[0005] However, the "peak loss method" has the following technical drawbacks:

[0006] 1. High computational cost: To accurately capture the peak loss, numerical algorithms need to perform additional local scans or iterative searches at each step point to locate the maximum value of a function. This process significantly increases the total computation time of the entire design, especially for complex waveguide structures, where the design process may take hours or even days.

[0007] 2. Numerical instability: The peak value of the mode coupling loss is highly sensitive to simulation parameters (such as mesh accuracy and step size) and is easily affected by numerical noise. This may lead to poor convergence of the feedback control algorithm, or even design failure in some cases, making the design process less robust.

[0008] Therefore, there is an urgent need for a new design method that can significantly improve computational efficiency and algorithm stability while ensuring that the design performance is close to that of the "peak loss method". Summary of the Invention

[0009] This invention aims to address the problems of high computational complexity and poor numerical stability in existing optical waveguide adiabatic tapered structure design methods, and provides a design method for optical waveguide adiabatic tapered structures based on mode coupling loss slope control. This invention no longer controls the peak value of mode coupling loss, but rather controls the initial rate of change of loss with propagation distance (i.e., the slope), keeping it constant throughout the tapering process. The physical basis for this is that the initial slope of the loss curve directly reflects the intensity of mode coupling and is a strongly correlated quantity and effective "proxy" for the magnitude of the loss peak. By controlling this computationally simpler slope, effective suppression of the loss peak can be indirectly achieved.

[0010] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0011] The design method for adiabatic tapered structures of optical waveguides based on mode coupling loss slope control includes the following steps:

[0012] Step 1: Initialization: Define the initial geometric parameters of the tapered waveguide (e.g., input width W). in Termination geometry parameters (such as output width W) out ), material refractive index, operating wavelength λ, etc. A target constant slope value S is set. target This value represents the permissible level of mode coupling intensity and is a small positive number. An initial propagation step size Δz is set. The current position z is initialized to 0, and the current waveguide width W(z) is initialized to W... in Step 2: Calculate the local loss slope; Step 3: Adjust the local cone angle; Step 4: Iterate forward; Step 5: Generate the final structure.

[0013] As a further preferred embodiment of the present invention, in step two, at the current position z, it is assumed that the waveguide width extends forward with a test local cone angle α(z); using a numerical mode solver (such as the finite difference method FDE, the eigenmode expansion method EME, etc.), the coupling power loss P between the fundamental mode and the higher-order mode within a small distance from z to z+δz is calculated. loss (z′), where z′∈[z, z+δz];

[0014] Calculate the initial slope S of the loss curve at the starting point z. local :

[0015]

[0016] In the numerical implementation, the calculation is performed using a difference approximation: ;

[0017] This calculation does not require finding the peak value; it only requires a small forward calculation step.

[0018] Furthermore, as a preferred embodiment of the present invention, in step three, the calculated local slope S is... local With the preset target slope S target Compare them.

[0019] A feedback control algorithm (such as a proportional-integral-derivative PID controller, or a simple proportional controller) is applied to adjust the local cone length function L(z), thereby determining the actual cone angle α(z) at the current position. The adjustment logic is as follows:

[0020] If S local > S target This indicates that the current cone change is too rapid, and the mode coupling is too strong. It is necessary to reduce the local cone angle, that is, to make the cone change more gradual.

[0021] If S local < S target This indicates that the current cone transition is too gradual and there is room for acceleration. The local cone angle can be increased, making the cone steeper, to shorten the overall length.

[0022] The goal of this adjustment process is to make S local ≈ S target .

[0023] As a further preferred embodiment of the present invention, step four calculates the width W(z+Δz) of the waveguide profile at the next step point z+Δz based on the local cone angle α(z) that meets the conditions determined in step three; updates the current position z←z+Δz; and repeats steps two to four until the current waveguide width W(z) reaches the target output width W. out .

[0024] Furthermore, as a preferred embodiment of the present invention, step five connects all the calculated coordinate points (z, W(z)) to form the final and optimized tapered waveguide profile.

[0025] The optical waveguide adiabatic tapered structure design method based on mode coupling loss slope control described in this invention has the following technical advantages compared with existing technologies:

[0026] (1) Extremely high computational efficiency: The core of this method is to calculate the initial slope of the loss curve, which is a non-iterative, non-search calculation process. Its computational complexity is much lower than that of finding the loss peak. Under the same hardware conditions, the design speed of this invention can be increased by more than an order of magnitude compared with the "peak loss method", which greatly shortens the device development cycle.

[0027] (2) Strong algorithm robustness: The slope of a function is usually less sensitive to simulation parameters such as numerical noise and mesh accuracy than its peak value. Therefore, the slope-based feedback control loop is more stable, has better convergence, and improves the design success rate and the reliability of the results.

[0028] (3) Excellent performance: Although this method does not directly control the peak loss, due to the strong physical positive correlation between the initial slope and the peak value, the tapered structure designed by this method is very close to the structure designed by the "peak loss method" with high computational cost in terms of key performance indicators such as transmission loss and device length, thus achieving the best balance between performance and efficiency.

[0029] (4) Good universality: This method is applicable to the design of tapered structures for various material systems (such as silicon-based, indium phosphide-based, silicon nitride-based, etc.) and various waveguide types (such as strip waveguide, ridge waveguide, slot waveguide, etc.). Attached Figure Description

[0030] Figure 1 This is a flowchart of the method of the present invention in an embodiment of the present invention. Detailed Implementation

[0031] The present invention will be further explained in detail below with reference to the accompanying drawings, so that those skilled in the art can better understand and implement the present invention. However, the following examples are only used to explain the present invention and are not intended to limit the present invention.

[0032] like Figure 1 As shown, the design method for an adiabatic tapered structure of an optical waveguide based on mode coupling loss slope control includes the following steps:

[0033] Step 1: Initialization: Define the initial geometric parameters of the tapered waveguide (e.g., input width W). in Termination geometry parameters (such as output width W) out ), material refractive index, operating wavelength λ, etc. A target constant slope value S is set. target This value represents the permissible level of mode coupling intensity and is a small positive number. An initial propagation step size Δz is set. The current position z is initialized to 0, and the current waveguide width W(z) is initialized to W... in Step 2: Calculate the local loss slope; Step 3: Adjust the local cone angle; Step 4: Iterate forward; Step 5: Generate the final structure.

[0034] In step two, at the current position z, assume the waveguide width extends forward with a test local cone angle α(z); using a numerical mode solver (such as the finite difference method FDE, the eigenmode expansion method EME, etc.), calculate the coupling power loss P between the fundamental mode and the higher-order mode over a small distance from z to z+δz. loss(z′), where z′∈[z, z+δz];

[0035] Calculate the initial slope S of the loss curve at the starting point z. local :

[0036]

[0037] In the numerical implementation, the calculation is performed using a difference approximation: ;

[0038] This calculation does not require finding the peak value; it only requires a small forward calculation step.

[0039] In step three, the calculated local slope S local With the preset target slope S target Compare them.

[0040] A feedback control algorithm (such as a proportional-integral-derivative PID controller, or a simple proportional controller) is applied to adjust the local cone length function L(z), thereby determining the actual cone angle α(z) at the current position. The adjustment logic is as follows:

[0041] If S local > S target This indicates that the current cone change is too rapid, and the mode coupling is too strong. It is necessary to reduce the local cone angle, that is, to make the cone change more gradual.

[0042] If S local < S target This indicates that the current cone transition is too gradual and there is room for acceleration. The local cone angle can be increased, making the cone steeper, to shorten the overall length.

[0043] The goal of this adjustment process is to make S local ≈ S target .

[0044] Step 4: Based on the local cone angle α(z) determined in Step 3 that meets the conditions, calculate the width W(z+Δz) of the waveguide profile at the next step point z+Δz; update the current position z←z+Δz; repeat steps 2 to 4 until the current waveguide width W(z) reaches the target output width W. out .

[0045] Step 5 connects all the calculated coordinate points (z, W(z)) to form the final, optimized tapered waveguide profile.

[0046] In a specific implementation, take a ridge waveguide to strip waveguide mode converter on an SOI (silicon-on-insulator) platform as an example;

[0047] Device parameters: Operating wavelength λ = 1550 nm. Top silicon thickness H = 220 nm, etching depth h = 130 nm.

[0048] Input: Ridge waveguide, width W in = 1.2 μm.

[0049] Output: Strip waveguide, width W out = 0.5 μm.

[0050] Target slope: S target = 1 × 10 −5 μm −1 (These are sample values ​​and can be adjusted according to performance requirements);

[0051] Implementation process:

[0052] (1) At z = 0, the waveguide width W(0) = 1.2 μm.

[0053] (2) Using commercial or self-developed model solving software (such as Lumerical MODE, COMSOL, etc.), establish a 2D model of the cross section and solve for the effective refractive index n of the fundamental mode (TE0) and the first higher-order mode (TE1). eff,0 and n eff,1 .

[0054] (3) Assume a test local tapering length L test Then the local cone angle α ≈ (W out -W in ) / L test .

[0055] (4) Mode coupling loss P loss This is mainly caused by the interference of these two modes, and the amplitude of the oscillation is proportional to the coupling coefficient κ and the beat length. The relationship between the initial slope S of the loss. local It is related to the local cone angle α and the mode parameters.

[0056] (5) Calculate the local slope S at point z. local This can be obtained by performing mode expansion calculations at point z and point z+δz (δz is a minimum value, such as 0.01 μm), obtaining the power loss at the two points, and then performing differential calculations.

[0057] (6) S local With S target Comparison. Assuming a proportional controller is used, the new local tapered length L new (z) can be accessed via L new (z) = L old (z)⋅(Slocal / S target The updated L(z) defines the true steepness of the cone at point z.

[0058] (7) Based on the updated L(z), determine the waveguide width W(z+Δz) at the next position z+Δz.

[0059] (8) Repeat this process until W(z) reaches 0.5 μm. The final point set (z, W(z)) constitutes a high-performance tapered profile.

[0060] Through the above embodiments, a short and low-loss tapered waveguide can be generated quickly and stably, with performance comparable to that obtained by the time-consuming "peak loss method", but with a design time reduced by more than 90%.

[0061] The specific implementation schemes described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific implementation schemes of the present invention and are not intended to limit the scope of the present invention. Any equivalent changes and modifications made by those skilled in the art without departing from the concept and principles of the present invention should fall within the scope of protection of the present invention.

Claims

1. A design method for an adiabatic tapered structure of an optical waveguide based on mode coupling loss slope control, characterized in that, Includes the following steps: Step 1: Initialization: Define the initial geometric parameters, termination geometric parameters, material refractive index, and operating wavelength λ of the tapered waveguide; Set a target constant slope value S target ; Set an initial propagation step size Δz; initialize the current position z to 0, and initialize the current waveguide width W(z) to W. in ; Step 2: Calculation of local loss slope; Step 3: Adjust the local cone angle; Step 4: Iterate forward; Step 5: Generate the final structure.

2. The design method for an optical waveguide adiabatic tapered structure based on mode coupling loss slope control according to claim 1, characterized in that, In step two, at the current position z, it is assumed that the waveguide width extends forward with a test local cone angle α(z); using a numerical mode solver, the coupling power loss P between the fundamental mode and the higher-order mode over a small distance from z to z+δz is calculated. loss (z′), where z′∈[z, z+δz]; Calculate the initial slope S of the loss curve at the starting point z. local : ; In the numerical implementation, the calculation is performed using a difference approximation: .

3. The design method for an optical waveguide adiabatic tapered structure based on mode coupling loss slope control according to claim 2, characterized in that, In step three, the calculated local slope S local With the preset target slope S target A comparison is made; a feedback control algorithm is applied to adjust the local cone length function L(z) to determine the actual cone angle α(z) at the current position. The adjustment logic is as follows: if S local > S target This indicates that the current cone change is too rapid, and the mode coupling is too strong; it is necessary to reduce the local cone angle, that is, to make the cone change more gradual; if S local < S target This indicates that the current cone angle is too gentle and there is room for acceleration; the local cone angle can be increased, i.e., the cone angle can be made steeper, in order to shorten the overall length; the goal of this adjustment process is to make S... local ≈ S target .

4. The design method for an optical waveguide adiabatic tapered structure based on mode coupling loss slope control according to claim 3, characterized in that, Step four calculates the waveguide profile width W(z+Δz) at the next step point z+Δz based on the local cone angle α(z) that meets the conditions determined in step three; updates the current position z←z+Δz; and repeats steps two to four until the current waveguide width W(z) reaches the target output width W. out .

5. The design method for an optical waveguide adiabatic tapered structure based on mode coupling loss slope control according to claim 4, characterized in that, Step five connects all the calculated coordinate points (z, W(z)) to form the final, optimized tapered waveguide profile.