WO2020135860A1 - Method for optimizing waveguide and cruciform waveguide crosser - Google Patents

Abstract

A method for optimizing a waveguide and a cruciform waveguide crosser. The method for optimizing a waveguide comprises: splitting a quasiadiabatic waveguide taper having a gradually changing width into N straight waveguides and setting a shape model for each straight waveguide (S110), wherein the shape model of the ith straight waveguide is Li=f(Wi), 1≤i≤N, Li is the length of the ith straight waveguide in a waveguide propagation direction, Wi is the width of the cross section of the ith straight waveguide, and the cross section is perpendicular to the waveguide propagation direction; determining, according to a transmission matrix of each straight waveguide and a coupling transmission matrix between two adjacent straight waveguides, a transmission matrix T of the quasiadiabatic waveguide taper having a target length (S120); optimizing, according to an insertion loss index of the quasiadiabatic waveguide taper, a parameter of the shape model of the quasiadiabatic waveguide taper (S130); and determining the shape of the quasiadiabatic waveguide taper according to the shape model and the optimized parameter (S140).

Description

Method for optimizing waveguide and cross waveguide crossover

Cross-reference of related applications

This application is based on a Chinese patent application with an application number of 201811639879.9 and an application date of December 29, 2018, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is hereby incorporated by reference.

Technical field

The invention relates to the field of optoelectronic technology, in particular to a method for optimizing waveguides and cross waveguide crossovers.

Background technique

In recent years, fast, low-cost and highly integrated silicon-based photonic integrated chips have been vigorously developed in technology, and gradually become commercially available. Silicon-based photonic integrated chips are widely used, such as optical network switches. The optical network switch can complete the function of switching the optical signal without switching to the digital field, and has great application prospects, such as all-optical networks, data centers, and optical interconnections. Silicon photonic switches are composed of optical waveguides of different lengths, multiple optical switch units, and multiple waveguide cross structures. The waveguide cross-over structure with lower insertion loss and lower crosstalk is essential.

Commonly used waveguide cross structures include multi-layer waveguide cross structures and single-layer waveguide cross structures. The multi-layer waveguide cross structure has extremely low crosstalk and insertion loss, but the preparation process is complicated. The single-layer waveguide cross structure is composed of a waveguide cross (Crossing) and a waveguide taper (Taper). The cross structure waveguide is often wider than the single-mode waveguide to reduce insertion loss.

In the single-layer waveguide crossing structure, the waveguide crossing optimized according to the principle of Multimode Interference (MMI) has the characteristics of low insertion loss and small size, but the process tolerance is also small and has a certain wavelength dependence; in order to reduce the waveguide For the insertion loss of the cross, a wide-width waveguide cross region can be used. In order to reduce the fundamental mode propagation insertion loss of the tapered region (Taper) region, a very long linear or exponential waveguide tapered taper is generally used. Small insertion loss and large process tolerance are achieved, but the size is large.

Summary of the invention

Embodiments of the present invention provide a method for optimizing a waveguide and a cross-waveguide crossover. Compared with related technologies, the waveguide size and insertion loss can be reduced.

An embodiment of the present invention provides a method for optimizing a waveguide, including:

The tapered section of the quasi-adiabatic waveguide with gradually varying width is decomposed into N sections of straight waveguides, and the shape model of each section of straight waveguide is set; wherein, the shape model of the section i straight waveguide is L _i =f(W _i ), 1≤i≤N , L _i is the length of the i-th straight waveguide in the waveguide propagation direction, and W _i is the width of the cross section of the i-th straight waveguide, which is perpendicular to the waveguide propagation direction;

The transmission matrix T of the quasi-adiabatic waveguide segment of the target length is determined according to the transmission matrix of each straight waveguide and the coupling transmission matrix between two adjacent straight waveguides;

Optimize the parameters of the shape model of the quasi-insulated waveguide cone section according to the insertion loss index of the quasi-insulated waveguide cone section;

The shape of the quasi-insulated waveguide cone section is determined according to the shape model and the optimized parameters.

An embodiment of the present invention provides a cross waveguide cross-connect, including:

A first waveguide and a second waveguide with the same structure, the first waveguide and the second waveguide cross vertically at the center;

The first waveguide includes a first input quasi-insulated waveguide cone section, a first cross-region waveguide section and a first output quasi-insulated waveguide cone section, and the first input quasi-insulated waveguide cone section includes a narrow mouth and a wide mouth and passes The wide opening is connected to the first end of the first cross-sectional waveguide section, and the first output quasi-insulated waveguide cone section includes a narrow port and a wide opening and passes through the wide opening to the first cross-sectional waveguide The second end of the segment is connected;

The second waveguide includes a second input quasi-insulated waveguide cone section, a second cross-region waveguide section and a second output quasi-insulated waveguide cone section, and the second input quasi-insulated waveguide cone section includes a narrow mouth and a wide mouth and passes The wide opening is connected to the third end of the second cross-sectional waveguide section, and the second output quasi-insulated waveguide cone section includes a narrow opening and a wide opening and passes through the wide opening to the second cross-sectional waveguide The fourth end of the segment is connected;

The shapes of the first input quasi-adiabatic waveguide cone section, the first output quasi-adiabatic waveguide cone section, the second input quasi-adiabatic waveguide cone section and the second output quasi-adiabatic waveguide cone section are all determined by the method of optimizing the waveguide described above.

Other features and corresponding beneficial effects of the present invention are explained in the later part of the description, and it should be understood that at least part of the beneficial effects will become apparent from the description in the description of the present invention.

BRIEF DESCRIPTION

1 is a flowchart of a method for optimizing a waveguide according to Embodiment 1 of the present invention;

Figure 2-a is a schematic diagram of a decomposition of a quasi-insulated waveguide cone section in an embodiment of the present invention;

Figure 2-b is a schematic diagram of another decomposition of the quasi-insulated waveguide cone section in the embodiment of the present invention;

3 is a schematic diagram of a cross waveguide crossover according to Embodiment 2 of the present invention;

4 is a diagram of the relationship between the width of the cross region of the waveguide and the insertion loss in an embodiment of the present invention;

Figure 5-a is a schematic diagram of a ridge waveguide in Example 1 of the present invention;

Fig. 5-b is a schematic diagram of the simulation of the relationship between the parameters of the width and length of the optimized quasi-insulated waveguide taper (Taper) in Example 1 of the present invention;

5-c is a schematic diagram of the simulation of the fundamental mode insertion loss of the optimized quasi-insulated waveguide cone section in Example 1 of the present invention;

Fig. 5-d is a schematic diagram of the fundamental mode insertion loss simulation of the cross-waveguide crossover (WC 1) based on the optimized quasi-insulated waveguide cone section in Example 1 of the present invention;

6-a is a schematic diagram of the simulation of the relationship between the parameters of the width and length of the optimized quasi-insulated waveguide taper (Taper) in Example 2 of the present invention;

6-b is a schematic diagram of the simulation of the fundamental mode insertion loss of the optimized quasi-insulated waveguide cone section in Example 2 of the present invention;

6-c is a schematic diagram of the simulation of the fundamental mode insertion loss of the cross waveguide crossover (WC 1) based on the optimized quasi-insulated waveguide cone section in Example 2 of the present invention;

7-a is a schematic diagram of the simulation of the relationship between the parameters of the width and length of the optimized quasi-insulated waveguide taper (Taper) in Example 3 of the present invention;

7-b is a schematic diagram of simulation of the fundamental mode insertion loss of the optimized quasi-insulated waveguide cone section in Example 3 of the present invention;

7-c is a schematic diagram of a simulation of the fundamental mode insertion loss of a cross-waveguide crossover (WC 1) based on an optimized quasi-insulated waveguide cone section in Example 3 of the present invention;

8-a is a schematic diagram of simulation of the relationship between parameters of the width and length of the optimized quasi-insulated waveguide taper (Taper) in Example 4 of the present invention;

8-b is a schematic diagram of the simulation of the fundamental mode insertion loss of the optimized quasi-insulated waveguide cone section in Example 4 of the present invention;

8-c is a schematic diagram of the simulation of the fundamental mode insertion loss of the cross waveguide crossover (WC 1) based on the optimized quasi-insulated waveguide cone section in Example 4 of the present invention.

detailed description

To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that the embodiments in the present application and the features in the embodiments can be arbitrarily combined with each other without conflict.

Quasi-adiabatic waveguide refers to a waveguide whose structure changes slowly and the propagation mode does not undergo useless mode conversion. In the related art, quasi-adiabatic waveguides often use very long tapers to meet quasi-adiabatic conditions. The application realizes the rapid insulation effect of the quasi-insulated waveguide, and the mode conversion characteristics are used to change the shape of the cone section, so that the useless mode conversion in mode propagation is reduced, so that the quasi-insulation condition is satisfied on the shorter length of the cone section.

This application uses the transmission matrix method to quickly complete the optimization of the shape of the quasi-adiabatic waveguide cone section with multiple modes, to maximize the retention of TE0 fundamental mode component propagation, and to reduce insertion loss.

Example 1

As shown in FIG. 1, an embodiment of the present invention provides a method for optimizing a waveguide, including:

In step S110, the tapered section of the quasi-adiabatic waveguide with gradually varying width is decomposed into N sections of straight waveguides, and the shape model of each section of straight waveguide is set; wherein, the shape model of the section i straight waveguide is L _i =f(W _i ), 1≤ i≤N, L _i is the length of the segment i of the straight waveguide in the propagation direction of the waveguide, the cross-sectional width W is _i is the i-th straight waveguide section, the cross section perpendicular to the propagation direction of the waveguide;

Step S120: Determine the transmission matrix T of the quasi-adiabatic waveguide cone segment of the target length according to the transmission matrix of each straight waveguide and the coupling transmission matrix between two adjacent straight waveguides;

Step S130: Optimize the parameters of the shape model of the quasi-insulated waveguide cone section according to the insertion loss index of the quasi-insulated waveguide cone section;

Step S140: Determine the shape of the quasi-insulated waveguide cone section according to the shape model and the optimized parameters.

In one embodiment, the shape model includes the following first model or second model:

The first model:

The second model:

Among them, A ₁ ～A ₉ are the parameters of the first model; A ₁ ～A ₁₅ are the parameters of the second model;

Among them, as shown in Figure 2-a and Figure 2-b, there are two ways to decompose the quasi-insulated waveguide cone. In the decomposition mode shown in Fig. 2-a, the width of the first straight waveguide is equal to the width of the first port of the tapered section of the quasi-insulated waveguide. In the decomposition mode shown in Fig. 2-b, the width of the last straight waveguide is equal to the width of the second port of the tapered section of the quasi-insulated waveguide. Two methods of decomposing the quasi-insulated waveguide cone are available.

In one embodiment, the determining the transmission matrix T of the quasi-adiabatic waveguide segment of the target length according to the transmission matrix of each straight waveguide and the coupling transmission matrix between two adjacent straight waveguides includes:

When the quasi-adiabatic tapered section of the target length includes continuous m-section straight waveguides, determine the transmission matrix of the m-section straight waveguide and the coupling transmission matrix between two adjacent sections of straight waveguides;

The transmission matrix T of the quasi-adiabatic waveguide cone of the target length is determined in the following way:

Where P _i is the transmission matrix of the i-th straight waveguide, T _i is the coupling transmission matrix between the i-th straight waveguide and the adjacent i+1-th straight waveguide, and ∏ is the multiplication symbol.

Among them, P _i and T _i are both n*n-dimensional matrices, n is the order of modes existing in the waveguide, and the larger the n, the more accurate the calculation of the transmission matrix.

Through the mode analysis and calculation of waveguides of different widths, the spatial distribution of the electric field magnetic field of each order mode can be obtained. By calculating the overlapping integral of the mode fields of adjacent width waveguides, the modulus square of each element value of the T _i matrix is obtained, according to the coupled mode theory to give each matrix element values T _i.

In the P _i matrix, the non-diagonal elements are zero, and the diagonal elements are respectively

k _m is the propagation constant of the m-th mode field in the waveguide, and L _i is the length of the i-th waveguide.

In one embodiment, the insertion loss index of the quasi-insulated waveguide cone segment includes:

The fundamental matrix insertion loss of the transmission matrix T of the overall quasi-adiabatic waveguide cone section is the smallest; where the overall quasi-adiabatic waveguide cone section includes all straight waveguides.

In one embodiment, the insertion loss index of the quasi-insulated waveguide cone segment includes:

The fundamental matrix insertion loss of the transmission matrix of all quasi-adiabatic waveguides in the local cone section is less than the insertion loss threshold, and the fundamental mode insertion loss of the transmission matrix of the overall quasi-adiabatic waveguide cone section is the smallest;

Among them, the j-th quasi-adiabatic waveguide partial cone segment refers to: a continuous j-segment straight waveguide from the first-segment straight waveguide to the end of the m- _th j-th straight waveguide.

The embodiment of the present invention uses the transmission matrix method to calculate the conversion between the modes in the tapered section (Taper). Practical research shows that this conversion is bidirectional, and the magnitude of higher-order mode components changes with the oscillation of the propagation direction, and the amplitude of the oscillation is affected by the shape of the taper. Therefore, by optimizing the shape of the cone segment, the fundamental mode component at the output end is maximized; at the same time, in order to achieve a high tolerance, the shape of the cone segment with a small amplitude variation of the mode component can be selected.

Example 2

As shown in FIG. 3, an embodiment of the present invention provides a cross waveguide cross-connector, including: a first waveguide 1 and a second waveguide 2 having the same structure, the first waveguide 1 and the second waveguide 2 perpendicularly cross at the center;

The first waveguide 1 includes a first input quasi-insulated waveguide cone section 11, a first cross-region waveguide section 12 and a first output quasi-insulated waveguide cone section 13, the first input quasi-insulated waveguide cone section includes a narrow mouth and A wide mouth is connected to the first end of the first cross-sectional waveguide segment through the wide mouth. The first output quasi-insulated waveguide cone section includes a narrow mouth and a wide mouth and passes through the wide mouth and the first The second end of a cross-section waveguide segment is connected;

The second waveguide 2 includes a second input quasi-adiabatic waveguide cone section 21, a second crossing region waveguide section 22 and a second output quasi-adiabatic waveguide cone section 23, and the second input quasi-adiabatic waveguide cone section includes a narrow mouth and A wide mouth is connected to the third end of the waveguide section of the second intersection region through the wide mouth, and the second output quasi-insulated waveguide cone section includes a narrow mouth and a wide mouth and passes through the wide mouth and the first The fourth end of the waveguide section in the two-cross region is connected

The shapes of the first input quasi-adiabatic waveguide cone section, the first output quasi-adiabatic waveguide cone section, the second input quasi-adiabatic waveguide cone section and the second output quasi-adiabatic waveguide cone section are all determined by the method of optimizing the waveguide described above;

The insertion loss of the cross waveguide crossover comes from the tapered section (Taper) and the cross section of the waveguide (Crossing). Fig. 4 shows the relationship between the width of the cross region of the waveguide and the insertion loss of the fundamental mode of the region, where the horizontal axis (x) is the width of the cross region, the unit is micrometer (um), and the vertical axis (TE0IL) is The insertion loss of the fundamental mode in the cross region is in dB. The wider the waveguide cross region is, the smaller the fundamental mode propagation loss is, and the change is more and more gentle. When it is greater than 4um, the insertion loss of the waveguide cross region is <0.1dB. Using a wider waveguide cross region can reduce the insertion loss of the waveguide cross, but a wider waveguide cross requires a longer waveguide cone section.

The insertion loss of the fundamental mode of the taper mainly comes from the non-adiabatic propagation of the mode between the input and output single-mode waveguide and the wide waveguide in the cross section of the waveguide, so that the conversion of the fundamental mode to the higher-order mode occurs. The embodiment of the present invention adopts a fast insulating waveguide cone section, thereby greatly reducing the mode conversion loss in the tape section (Taper) and shortening the length of the cone section.

In one embodiment, the first input quasi-insulated waveguide cone section, the first output quasi-insulated waveguide cone section, the second input quasi-insulated waveguide cone section, and the second output quasi-insulated waveguide cone section have a wide mouth greater than or equal to 4 microns.

In one embodiment, the cross waveguide crossover has rotational symmetry: the cross waveguide crossover is rotated 90 degrees in a plane composed of the propagation direction of the first waveguide and the propagation direction of the second waveguide Coincides with before rotation.

In one embodiment, in the plane formed by the propagation directions of the first waveguide and the second waveguide, the optical field of the intersection area is not constrained in a direction perpendicular to the propagation direction of the optical path.

In one embodiment, both the first waveguide and the second waveguide are: a ridge waveguide; or the first waveguide and the second waveguide are: a strip waveguide.

The cross waveguide crossover of the present application and the method of optimizing the waveguide are explained below by examples.

The cross waveguide interleavers provided in Examples 1 to 4 below include a first waveguide and a second waveguide having the same structure, and the first waveguide and the second waveguide intersect perpendicularly at the center. The first waveguide includes a first input quasi-adiabatic waveguide cone segment, a first cross-region waveguide segment and a first output quasi-adiabatic waveguide cone segment, a second waveguide includes a second input quasi-adiabatic waveguide cone segment, and a second cross-region waveguide segment And the second output quasi-insulated waveguide cone section; the first input quasi-insulated waveguide cone section gradually changes from the first width to the second width, and the width of the waveguide section in the first intersection region is the second width, the first The output quasi-insulated waveguide tapered section gradually changes from the second width to the first width; the second input quasi-insulated waveguide tapered section gradually changes from the first width to the second width, and the width of the second cross-sectional waveguide section is the second Width, the second output quasi-insulated waveguide tapered from the second width to the first width. The cross waveguide crossover meets the requirement of rotational symmetry: the cross waveguide crossover coincides with the pre-rotation angle after being rotated 90 degrees in the plane formed by the propagation directions of the first waveguide and the second waveguide.

Example 1

In this example, the width of the intersection region is 6 microns, and the lengths of the first input quasi-insulated waveguide cone section, the first output quasi-insulated waveguide cone section, the second input quasi-insulated waveguide cone section, and the second output quasi-insulated waveguide cone section are all It is 100 microns, and the narrow mouth of each quasi-insulated waveguide cone is 0.5 microns wide and 6 microns wide.

As shown in Figure 5-a, both the first waveguide and the second waveguide are symmetrical shallow etched silicon waveguides. 131 is the ridge waveguide area, 132 is the shallow etched waveguide area, the two sides of the shallow etched area are symmetrical, the total width is 8um wider than the ridge waveguide, 4um on each side, the difference between the height of the ridge waveguide and the shallow etched area is the etched depth, The etching depth is 70 nm, the waveguide height is 220 nm, the input and output waveguide width is 0.5 μm, the cross section of the waveguide is 6 μm wide, and the upper and lower claddings are silicon oxide.

The shape model of each quasi-adiabatic waveguide cone section in this example uses the function of the first model:

The insertion loss index of the transmission matrix is used to optimize the shape model parameters of the adiabatic waveguide cone. The cone section of the quasi-adiabatic waveguide is decomposed into a combination of multiple straight waveguides, the width difference between two adjacent straight waveguides is 0.05um, and the transmission matrix T of the entire quasi-adiabatic waveguide cone section is determined according to the transmission matrix of each straight waveguide;

When the insertion loss of the transmission matrix T of the entire quasi-adiabatic waveguide cone section is required to be minimum and the insertion loss of any quasi-adiabatic waveguide cone section is less than or equal to the first threshold value, the extremely low fundamental mode insertion loss and Large structural tolerance. Among them, the j-th quasi-adiabatic waveguide partial cone segment refers to: a continuous j-segment straight waveguide from the first-segment straight waveguide to the end of the m- _th j-th straight waveguide.

By optimizing the parameters of the shape model, the values of A ₁ to A ₉ can be obtained as shown in Table 1 below:

A 1	A 2	A 3	A 4	A 5	A 6	A 7	A 8	A 9
-2.108	0.9334	10.18	52.79	2.640	3.529	0.9871	5.321	0.8861

Table 1

In Figure 5-b, after optimizing the shape of the tapered section of the quasi-insulated waveguide, the relationship between the width and length of the tapered section (Parameters) is shown in the figure, where the horizontal axis is the tapered section (Taper) ) Length (x), the unit is micrometer (um), the vertical axis is the width of the tapered section (Taper) (Width), the unit is micrometer (um).

In Figure 5-c, after optimizing the shape of the quasi-adiabatic waveguide cone (Taper), the simulation result of the optimized quasi-adiabatic waveguide cone's fundamental mode insertion loss is shown in the figure, where the horizontal axis is Wavelength and the unit is In micrometers (um), the vertical axis is the basic mode insertion loss (Insertion) of the tapered section (Taper). The optimized quasi-adiabatic waveguide taper has a fundamental mode insertion loss of <0.0045dB.

In Figure 5-d, after optimizing the shape of the quasi-adiabatic waveguide taper (Taper), the simulation results of the fundamental mode insertion loss of the cross-waveguide crossover (WC1) based on the optimized quasi-adiabatic waveguide taper are shown in the figure, where, The horizontal axis is Wavelength, the unit is micrometer (um), and the vertical axis is the insertion loss (Insertion) of the fundamental mode of the cross waveguide crossover (WC 1). Based on the optimized quasi-adiabatic tapered section of the cross waveguide cross-section (WC) 1 basic mode insertion loss <0.067dB.

Example 2

In this example, the width of the intersection area is 8 microns, and the lengths of the first input quasi-insulated waveguide cone section, the first output quasi-insulated waveguide cone section, the second input quasi-insulated waveguide cone section, and the second output quasi-insulated waveguide cone section are all It is 100 microns, and the narrow mouth of each quasi-insulated waveguide cone has a width of 0.5 microns and a wide mouth of 8 microns.

Both the first waveguide and the second waveguide are symmetrical shallow etched silicon waveguides with an etch depth of 70 nm, a waveguide height of 220 nm, a width of 0.5 μm for input and output waveguides, a width of 6 μm for the cross section of the waveguide, and a total width of the shallow etched waveguide area greater than that of the ridge waveguide 8um wide, the upper and lower cladding is silicon oxide.

The shape model of each quasi-insulated waveguide cone section in this example uses the function of the second model:

The insertion loss index of the transmission matrix is used to optimize the shape model parameters of the adiabatic waveguide cone. The cone section of the quasi-adiabatic waveguide is decomposed into a combination of multiple straight waveguides, the width difference between two adjacent straight waveguides is 0.05um, and the transmission matrix T of the entire quasi-adiabatic waveguide cone section is determined according to the transmission matrix of each straight waveguide;

When the insertion loss of the transmission matrix T of the entire quasi-adiabatic waveguide cone section is required to be minimum and the insertion loss of any quasi-adiabatic waveguide cone section is less than or equal to the first threshold value, the extremely low fundamental mode insertion loss and Large structural tolerance. Among them, the j-th quasi-adiabatic waveguide partial cone segment refers to: a continuous j-segment straight waveguide from the first-segment straight waveguide to the end of the m- _th j-th straight waveguide.

By optimizing the parameters of the shape model, the values of A ₁ to A ₁₅ can be obtained as shown in Table 2 below:

A 1	A 2	A 3	A 4	A 5	A 6
-4.887	0.0597	-0.01578	-0.1355	-1.778	0.1123

A 7	A 8	A 9	A 10	A 11	A 12
1.344E-04	-0.02869	5.735	1.539	3.241	0.2750
A 13	A 14	A 15	A	A	A
0.2001	1.382	1.017	A	A	A

Table 2

In Figure 6-a, after optimizing the shape of the tapered section of the quasi-insulated waveguide, the relationship between the width and length of the tapered section (Parameters) is shown in the figure, where the horizontal axis is the tapered section (Taper) ) Length (x), the unit is micrometer (um), the vertical axis is the width of the tapered section (Taper) (Width), the unit is micrometer (um).

In Figure 6-b, after optimizing the shape of the quasi-adiabatic waveguide cone (Taper), the simulation results of the optimized quasi-adiabatic waveguide cone's fundamental mode insertion loss are shown in the figure, where the horizontal axis is Wavelength and the unit is In micrometers (um), the vertical axis is the basic mode insertion loss (Insertion) of the tapered section (Taper). The optimized quasi-adiabatic waveguide taper has a fundamental mode insertion loss <0.007dB.

In Figure 6-c, after optimizing the shape of the quasi-adiabatic waveguide taper (Taper), the simulation results of the fundamental mode insertion loss of the cross waveguide crossover (WC 2) based on the optimized quasi-adiabatic waveguide taper are shown in the figure, where, The horizontal axis is the wavelength (Wavelength), the unit is micrometer (um), and the vertical axis is the fundamental mode insertion loss (Insertion) of the cross-waveguide crossover (WC 2). Based on the optimized quasi-adiabatic waveguide cone section of the WC 2 (WC), the fundamental mode insertion loss is <0.047dB.

Example 3

In this example, the width of the intersection region is 6 microns, and the lengths of the first input quasi-insulated waveguide cone section, the first output quasi-insulated waveguide cone section, the second input quasi-insulated waveguide cone section, and the second output quasi-insulated waveguide cone section are all It is 150 microns, and the narrow mouth of each quasi-insulated waveguide cone is 0.5 microns wide and 6 microns wide.

Both the first waveguide and the second waveguide are symmetrical shallow etched silicon waveguides with an etch depth of 70 nm, a waveguide height of 220 nm, a width of 0.5 μm for input and output waveguides, a width of 6 μm for the cross section of the waveguide, and a total width of the shallow etched waveguide area greater than that of the ridge waveguide 8um wide, the upper and lower cladding is silicon oxide.

The shape model of each quasi-insulated waveguide cone section in this example uses the function of the second model:

The insertion loss index of the transmission matrix is used to optimize the shape model parameters of the adiabatic waveguide cone. The cone section of the quasi-adiabatic waveguide is decomposed into a combination of multiple straight waveguides, the width difference between two adjacent straight waveguides is 0.05um, and the transmission matrix T of the entire quasi-adiabatic waveguide cone section is determined according to the transmission matrix of each straight waveguide;

When the insertion loss of the transmission matrix T of the entire quasi-adiabatic waveguide cone section is required to be minimum and the insertion loss of any quasi-adiabatic waveguide cone section is less than or equal to the first threshold value, the extremely low fundamental mode insertion loss and Large structural tolerance. Among them, the j-th quasi-adiabatic waveguide partial cone segment refers to: a continuous j-segment straight waveguide from the first-segment straight waveguide to the end of the m- _th j-th straight waveguide.

By optimizing the parameters of the shape model, the values of A ₁ to A ₁₅ can be obtained as shown in Table 3 below:

A 1	A 2	A 3	A 4	A 5	A 6
-366.4	-0.08935	-0.02216	0.09714	0.9747	-0.03334
A 7	A 8	A 9	A 10	A 11	A 12
0.2563	7.957E-05	-1.594	1.380	3.617	-0.3801
A 13	A 14	A 15	A	A	A
0.3329	-0.1849	-31.75	A	A	A

table 3

In Figure 7-a, after optimizing the shape of the tapered section of the quasi-insulated waveguide, the relationship between the width and length of the tapered section (Parameters) is shown in the figure, where the horizontal axis is the tapered section (Taper) ) Length (x), the unit is micrometer (um), the vertical axis is the width of the tapered section (Taper) (Width), the unit is micrometer (um).

In Figure 7-b, after optimizing the shape of the quasi-adiabatic waveguide cone (Taper), the simulation results of the optimized quasi-adiabatic waveguide cone's fundamental mode insertion loss are shown in the figure, where the horizontal axis is Wavelength and the unit is In micrometers (um), the vertical axis is the basic mode insertion loss (Insertion) of the tapered section (Taper). The optimized quasi-adiabatic waveguide taper has a fundamental mode insertion loss <0.002dB.

In Fig. 7-c, after optimizing the shape of the quasi-adiabatic waveguide taper (Taper), the simulation results of the fundamental mode insertion loss of the cross waveguide crossover (WC3) based on the optimized quasi-adiabatic waveguide taper are shown in the figure, where, The horizontal axis is Wavelength, the unit is micrometer (um), and the vertical axis is the insertion loss (Insertion) of the fundamental mode of the cross waveguide crossover (WC 3). The insertion loss of the fundamental mode of the cross-waveguide cross-over device (WC3) based on the optimized quasi-adiabatic tapered section is <0.06dB.

Example 4

In this example, the width of the intersection area is 8 microns, and the lengths of the first input quasi-insulated waveguide cone section, the first output quasi-insulated waveguide cone section, the second input quasi-insulated waveguide cone section, and the second output quasi-insulated waveguide cone section are all It is 150 microns, and the narrow mouth of each quasi-insulated waveguide cone is 0.5 microns wide and 8 microns wide.

Both the first waveguide and the second waveguide are symmetrical shallow etched silicon waveguides with an etch depth of 70 nm, a waveguide height of 220 nm, a width of 0.5 μm for input and output waveguides, a width of 8 μm for the cross section of the waveguide, and a difference in width between adjacent two straight waveguides of 0.05 um, the upper and lower cladding is silicon oxide.

The shape model of each quasi-insulated waveguide cone section in this example uses the function of the second model:

The insertion loss index of the transmission matrix is used to optimize the shape model parameters of the adiabatic waveguide cone. The quasi-adiabatic waveguide cone segment is decomposed into a combination of multi-section straight waveguides, the width difference between two adjacent straight waveguides is 0.05um, and the transmission matrix T of the entire quasi-adiabatic waveguide cone section is determined according to the transmission matrix of each straight waveguide;

When the insertion loss of the transmission matrix T of the entire quasi-adiabatic waveguide cone section is required to be minimum and the insertion loss of any quasi-adiabatic waveguide cone section is less than or equal to the first threshold value, the extremely low fundamental mode insertion loss and Large structural tolerance. Among them, the j-th quasi-adiabatic waveguide partial cone segment refers to: a continuous j-segment straight waveguide from the first-segment straight waveguide to the end of the m- _th j-th straight waveguide.

By optimizing the parameters of the shape model, the values of A ₁ to A ₁₅ can be obtained as shown in Table 4 below:

A 1	A 2	A 3	A 4	A 5	A 6
-1224.6	-0.082440	-0.43349E	0.24088	0.68336	0.12010
A 7	A 8	A 9	A 10	A 11	A 12
-0.31558	0.17353	-3.1440	-1.6827	-1.2815	0.16654
A 13	A 14	A 15	A	A	A
-1.6067	2.3663	43.517	A	A	A

Table 4

In Figure 8-a, after optimizing the shape of the tapered section of the quasi-insulated waveguide, the relationship between the width and length of the tapered section (Parameters) is shown in the figure, where the horizontal axis is the tapered section (Taper) ) Length (x), the unit is micrometer (um), the vertical axis is the width of the tapered section (Taper) (Width), the unit is micrometer (um).

In Figure 8-b, after optimizing the shape of the quasi-adiabatic waveguide cone (Taper), the simulation results of the optimized quasi-adiabatic waveguide cone's fundamental mode insertion loss are shown in the figure, where the horizontal axis is Wavelength and the unit is In micrometers (um), the vertical axis is the basic mode insertion loss (Insertion) of the tapered section (Taper). The optimized quasi-adiabatic waveguide taper has a fundamental mode insertion loss <0.0018dB.

In Fig. 8-c, after optimizing the shape of the quasi-adiabatic waveguide taper (Taper), the simulation results of the fundamental mode insertion loss of the cross waveguide crossover (WC 4) based on the optimized quasi-adiabatic waveguide taper are shown in the figure, where, The horizontal axis is Wavelength, the unit is micrometer (um), and the vertical axis is the insertion loss (Insertion) of the fundamental mode of the cross waveguide crossover (WC 4). Based on the optimized quasi-adiabatic waveguide cone section of the WC 4 (WC), the fundamental mode insertion loss is <0.047dB.

Compared with the related art, an embodiment of the present invention provides a method for optimizing a waveguide and a cross waveguide crossover, which decomposes a tapered section of a quasi-adiabatic waveguide with a gradually varying width into an N-section straight waveguide, and sets the shape model of each section of the straight waveguide, according to The transmission matrix of each straight waveguide and the coupling transmission matrix between two adjacent straight waveguides determine the transmission matrix T of the quasi-adiabatic waveguide cone segment of the target length, and optimize the quasi-adiabatic waveguide cone according to the insertion loss index of the quasi-adiabatic waveguide cone segment The parameters of the shape model of the segment determine the shape of the cone section of the quasi-insulated waveguide according to the shape model and the optimized parameters. By optimizing the shape of the quasi-insulated waveguide cone section, the waveguide size and insertion loss can be reduced. For the same length and wide opening size, the insertion loss of the optimized shape of the quasi-insulated waveguide cone section is significantly reduced; for the wide opening size of the quasi-insulated waveguide cone section, the length of the optimized shape of the quasi-insulated waveguide cone section is significantly reduced.

Those of ordinary skill in the art may understand that all or some of the steps, systems, and functional modules/units in the method disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware implementation, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be composed of several physical The components are executed in cooperation. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit . Such software may be distributed on computer-readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As is well known to those of ordinary skill in the art, the term computer storage medium includes both volatile and nonvolatile implemented in any method or technology for storing information such as computer readable instructions, data structures, program modules, or other data Sex, removable and non-removable media. Computer storage media include but are not limited to RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, magnetic tape, magnetic disk storage or other magnetic storage devices, or may Any other medium used to store desired information and accessible by a computer. In addition, it is well known to those of ordinary skill in the art that the communication medium generally contains computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium .

It should be noted that the present invention may have various other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications according to the present invention, but these Corresponding changes and deformations should belong to the protection scope of the claims appended to the present invention.

Claims (10)

1. A method of optimizing waveguides includes:

   The tapered section of the quasi-adiabatic waveguide with gradually varying width is decomposed into N sections of straight waveguides, and the shape model of each section of straight waveguide is set; wherein, the shape model of the section i straight waveguide is L _i =f(W _i ), 1≤i≤N , L _i is the length of the i-th straight waveguide in the waveguide propagation direction, and W _i is the width of the cross section of the i-th straight waveguide, which is perpendicular to the waveguide propagation direction;

   The transmission matrix T of the quasi-adiabatic waveguide segment of the target length is determined according to the transmission matrix of each straight waveguide and the coupling transmission matrix between two adjacent straight waveguides;

   Optimize the parameters of the shape model of the quasi-insulated waveguide cone section according to the insertion loss index of the quasi-insulated waveguide cone section;

   The shape of the quasi-insulated waveguide cone section is determined according to the shape model and the optimized parameters.
2. The method of claim 1, wherein:

   The shape model includes the following first model or second model:

   The first model:

   

   The second model:

   

   Among them, A ₁ ～A ₉ are the parameters of the first model; A ₁ ～A ₁₅ are the parameters of the second model.
3. The method of claim 1, wherein:

   The determining the transmission matrix T of the quasi-adiabatic waveguide cone section of the target length according to the transmission matrix of each straight waveguide section and the coupling transmission matrix between two adjacent straight waveguide sections includes:

   When the quasi-adiabatic tapered section of the target length includes continuous m-section straight waveguides, determine the transmission matrix of the m-section straight waveguide and the coupling transmission matrix between two adjacent sections of straight waveguides;

   The transmission matrix T of the quasi-adiabatic waveguide cone of the target length is determined in the following way:

   

   Where, P _i is the transmission matrix of the i-th straight waveguide, and T _i is the coupling transmission matrix between the i-th straight waveguide and the adjacent i+1-th straight waveguide.
4. The method of claim 3, wherein:

   The insertion loss index of the quasi-adiabatic waveguide cone section includes:

   The fundamental matrix insertion loss of the transmission matrix T of the overall quasi-adiabatic waveguide cone section is the smallest; where the overall quasi-adiabatic waveguide cone section includes all straight waveguides.
5. The method of claim 3, wherein:

   The insertion loss index of the quasi-adiabatic waveguide cone section includes:

   The fundamental matrix insertion loss of the transmission matrix of all quasi-adiabatic waveguides in the local cone section is less than the insertion loss threshold, and the fundamental mode insertion loss of the transmission matrix of the overall quasi-adiabatic waveguide cone section is the smallest;

   Among them, the j-th quasi-adiabatic waveguide partial cone segment refers to: a continuous j-segment straight waveguide from the first-segment straight waveguide to the end of the m- _th j-th straight waveguide.
6. A cross waveguide crossover device, including:

   A first waveguide and a second waveguide with the same structure, the first waveguide and the second waveguide cross vertically at the center;

   The first waveguide includes a first input quasi-insulated waveguide cone section, a first cross-region waveguide section and a first output quasi-insulated waveguide cone section, and the first input quasi-insulated waveguide cone section includes a narrow mouth and a wide mouth and passes The wide opening is connected to the first end of the first cross-sectional waveguide section, and the first output quasi-insulated waveguide cone section includes a narrow port and a wide opening and passes through the wide opening to the first cross-sectional waveguide The second end of the segment is connected;

   The second waveguide includes a second input quasi-insulated waveguide cone section, a second cross-region waveguide section and a second output quasi-insulated waveguide cone section, and the second input quasi-insulated waveguide cone section includes a narrow mouth and a wide mouth and passes The wide opening is connected to the third end of the second cross-sectional waveguide section, and the second output quasi-insulated waveguide cone section includes a narrow opening and a wide opening and passes through the wide opening to the second cross-sectional waveguide The fourth end of the segment is connected;

   The shapes of the first input quasi-adiabatic waveguide cone section, the first output quasi-adiabatic waveguide cone section, the second input quasi-adiabatic waveguide cone section and the second output quasi-adiabatic waveguide cone section are defined by any of claims 1-5 above. It is determined by the method for optimizing the waveguide.
7. The cross waveguide crossbar according to claim 6, wherein:

   The first input quasi-insulated waveguide cone section, the first output quasi-insulated waveguide cone section, the second input quasi-insulated waveguide cone section, and the second output quasi-insulated waveguide cone section have wide openings greater than or equal to 4 microns.
8. The cross waveguide crossbar according to claim 6, wherein:

   The cross waveguide crossover has a rotational symmetry: the cross waveguide crossover coincides with the pre-rotation angle after being rotated 90 degrees in a plane composed of the propagation direction of the first waveguide and the propagation direction of the second waveguide.
9. The cross waveguide crossbar according to claim 6, wherein:

   In the plane formed by the propagation directions of the first waveguide and the second waveguide, the optical field of the intersection area is not constrained in the direction perpendicular to the propagation direction of the optical path.
10. The cross waveguide crossbar according to claim 6, wherein:

    The first waveguide and the second waveguide are: ridge waveguide; or

    Both the first waveguide and the second waveguide are: strip waveguides.

Images