JPH01118822A - Optical matrix switch device - Google Patents

Abstract

PURPOSE:To perform switching at a low operating voltage without polarization dependency by constituting the optical matrix switch of a front stage by using N 1XN tree type optical switches and the optical matrix switch of a rear stage by using NX1 tree type optical switches. CONSTITUTION:The optical matrix switch 12 of the front stage is composed of the N 1XN tree type optical switches and the optical matrix switch 14 of the rear stage is composed of the N NX1 tree type optical switches. Then one light beam and the other light beams are inputted to the matrix switch 12 of the front stage, inputted to the matrix switch 16 of the rear stage from the matrix switch 12 of the front stage through a deviation varying means 16, and then outputted from the optical matrix switch 16 of the rear stage to the optical circuit element of a next stage. Consequently, the former and latter light beams are switched at the low operating voltage without any polarization characteristics.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention is directed to a polarization-independent optical matrix switch that electrically switches the transmission path of an optical signal.

(Prior Art) Optical matrix switches, which are constructed by providing waveguides and optical switches on a substrate, have advantages such as being able to be constructed in a compact size and operating voltage can be reduced, and research and development thereof is being actively conducted. ing.

This type of optical switch uses optical effects such as electro-optic effect and acousto-optic effect to switch signal light, but research into optical switches using electro-optic effect is currently underway as a practical optical device. It's progressing. This optical switch is constructed using a ferroelectric crystal substrate having an electro-optic effect, such as a LiNbO3 substrate. When an electric field is applied to an electro-optic crystal substrate via a control electrode, the refractive index of the substrate changes, so signal light can be switched using the change in refractive index. Changes in the refractive index are closely related to the direction P of the electric field applied to the substrate and the vibration direction (polarization direction) Q of the electric field of the signal light, and the electric field direction P becomes a certain direction and the polarization changes. When the direction Q is a specific direction, the refractive index can be efficiently changed with a low operating voltage (the electro-optic effect becomes large).However, when the electric field direction P and polarization direction Q are specified, As the refractive index deviates from the direction of Optical switches are configured so that they can switch light only when input signal light has a specific polarization direction.

Therefore, depending on the polarization direction of the input signal light, switching may or may not be performed, which is polarization dependence.

On the other hand, the literature I rJournal of Techno
lo9y (Journal Op Technology) J
Vol, LT-4, No.

II NOVEMBER1986pp, l717J l
c: Proposed Optical switches that do not have polarization dependence, such as the Saletail optical switch, are also being considered.In Bunjuuko's optical switch, a LiNbO3 substrate with the above-mentioned properties is used as the substrate. The configuration is such that a signal light having a specific polarization direction and a signal light having a polarization direction different from the specific polarization direction can be input and switching between these two signal lights can be performed.

Furthermore, devices that are constructed using a compound semiconductor substrate such as GaAs or InP and that switch signal light by utilizing the electro-optic effect have also been proposed. However, conventionally proposed optical switches that utilize the electro-optic effect require excessive operating voltage for light with a polarization direction different from a specific direction, making it virtually impossible to control switching, and this results in polarization dependence. I had it. Therefore, in order to eliminate polarization dependence, a method has been proposed in which the refractive index of a compound semiconductor substrate is changed by current injection. In this optical switch, the optical switch is operated in a total reflection state or a transmission state by changing the refractive index of a current injection region, thereby switching signal light.

(Problems to be Solved by the Invention) However, in the conventional optical switch without polarization dependence of the above-mentioned document I, the electro-optic effect is not limited to signal light having a specific polarization direction in which the electro-optic effect is large. In order to switch even signal lights with different polarization directions, the operating voltage is about 5 to 6 times higher than the operating voltage when switching only signal lights with a specific polarization direction. The problem was that it was necessary.

In addition, in optical switches that are not polarization-dependent and are constructed using compound semiconductor substrates, changes in the refractive index of the substrate are controlled by current injection, so the current consumption is lower than that of optical switches that utilize electro-optic effects. There were problems in that the number of optical switches increased and the operating speed of the optical switch slowed down.

The first object of the invention of this application is to solve the above-mentioned problems of conventional optical switches constructed using ferroelectric crystal substrates, and to perform polarization-independent switching at a lower operating voltage. The purpose of the present invention is to provide an optical matrix switch MIFr.

In addition, the second invention of this application solves the above-mentioned problems of an optical switch constructed using a compound semiconductor substrate, and provides a polarization-dependent structure that can reduce current consumption and increase operating speed. The purpose of this invention is to provide an optical matrix switch device M that is not available.

(Means for solving the problem) In order to achieve this object, the first invention of this application is constructed using a ferroelectric crystal substrate, and two lights having polarization directions orthogonal to each other are inputted. A boar-stage optical matrix switch constructed using a ferroelectric crystal substrate and outputting two lights having mutually orthogonal polarization directions, and these front-stage and rear-stage optical matrix switches. are connected in series, and bias changing means for changing the bias direction on the input surface of the subsequent optical matrix switch to a direction different from the polarization direction on the output surface of the preceding optical matrix switch, In an optical matrix switch device, the control optical switch in the first stage is a control optical switch that selects one of the light propagation paths, and the second stage optical matrix switch's control optical switch is a control optical switch that selects the other light propagation path. The optical matrix switch is constructed as follows. That is, the optical matrix switch at the front stage is composed of N 1×N tree-type optical switches, and the optical matrix switch at the rear stage IFr is composed of N NX1 tree-type optical switches.

Moreover, the second invention of this application is constructed using a compound semiconductor crystal substrate, and has two polarized directions perpendicular to each other.
a front-stage optical matrix switch into which two lights are input; a rear-stage optical matrix switch configured using a compound semiconductor crystal substrate and outputting two lights having mutually orthogonal polarization directions; Matrix switches are connected in series, and a bias changing means is provided for changing the polarization direction on the input surface of the subsequent optical matrix switch to a direction different from the polarization direction on the output surface of the preceding optical matrix switch. In an optical matrix switch device, the control optical switch of the matrix switch is a control optical switch that selects one propagation path of light, and the control optical switch of the subsequent optical matrix switch is a control optical switch that selects the propagation path of the other light. , the front-stage and rear-stage optical matrix switches are configured as follows. In other words, the front-stage optical matrix switch is composed of N 1×N tree-type optical switches, and the rear-stage optical matrix switch is composed of N 1×N tree-type optical switches.
It is constructed with an X1 tree type optical switch.

(Function) According to the optical matrix switch snowmaking of the first invention having such a configuration, one and the other light are input to the matrix switch at the front stage, and are transmitted from the matrix switch at the front stage to the matrix switch at the rear stage via the bias changing means. The signal is input to the matrix switch, and output from the optical matrix switch at the subsequent stage to the optical circuit element at the next stage.

Therefore, with one light having a polarization direction such that switching can be performed with a lower operating voltage, and with the other light having a polarization direction such that switching requires a higher operating voltage, Even if the input light is input to the optical matrix switch in the previous stage, by changing the polarization direction using the polarization changing means, the polarization direction is changed so that switching can be performed at a lower operating voltage, and the other light is sent to the optical matrix switch in the latter stage. You can input it.

As a result, the optical matrix switch at the front stage selectively changes the propagation path of one light at a low operating voltage), and the optical matrix switch at the rear stage selectively changes the propagation path of the other light at a low operating voltage. It can be selected (selectively changed).

Further, according to the optical matrix switch device of the second invention, one light and the other light are inputted to the previous stage matrix switch, and inputted from the previous stage matrix switch to the latter stage matrix switch via the bias changing means, Then, it is outputted from the optical matrix switch in the subsequent stage to the optical circuit element in the next stage.

Therefore, even if one light is input to the optical matrix switch in the previous stage with a polarization direction that can be switched and controlled by the electro-optic effect, and the other light is input in a polarization direction that cannot be substantially controlled by the electro-optic effect, The other light can be input to the subsequent matrix switch as the polarization direction can be controlled by switching by the electro-optic effect via the polarization changing means.

As a result, the optical matrix switch at the front stage selectively changes the propagation path of one light by switching it using the electro-optic effect), and the optical matrix switch at the rear stage changes the propagation path of the other light using the electro-optic effect. It is possible to select (selectively change) by switching.

That is, when configured using a compound semiconductor substrate, a switching operation without polarization dependence can be performed by utilizing the electro-optic effect.

(Embodiments) Hereinafter, embodiments of the first and second inventions of this application will be described with reference to the drawings. It should be noted that the drawings are only schematic illustrations to the extent that these inventions can be understood, and therefore, the dimensions, shapes, and distribution relationships of each component are not limited to the illustrated examples.

First Invention> Grave Two-Fork Stream (Schematic Description of Configuration) FIG. 1 is a plan view schematically showing the configuration of an embodiment of the first invention.

The optical matrix switch device flO of this embodiment is constructed using a ferroelectric crystal substrate 18a, and has a front-stage optical matrix switch 12 (hereinafter simply referred to as a front-stage switch) into which two lights having polarization directions orthogonal to each other are input. )and,
A rear-stage optical matrix switch 14 (hereinafter simply referred to as a rear-stage switch) configured using a ferroelectric crystal substrate + sb and outputting two lights having mutually orthogonal polarization directions.
and a bias changing means 16 that connects the front-stage switch 12 and the rear-stage switch 14 in series. The bias changing means 16 changes the bias direction on the input surface of the rear stage switch 14 to a direction different from the bias direction on the output surface of the front stage switch 12.

The front switch 10 has a substrate 18a, for example, 2-cut LiN.
It comprises a b○3 board and a suitable number (N) of 1×N tree-type optical switches 20 provided on this board 18 (N is a natural number).

The tree-type optical switch 20 has a tree-shaped propagation path that branches from one input port to N output ports. In this embodiment, in order to form this propagation path, a plurality of stages of controlled optical switches (hereinafter referred to as control switches) 22 are provided between the input and output ports of the tree-type optical switch 20. Each stage is provided with an arbitrary number of control switches 22. One input port and N output ports of the tree-type optical switch 20 are connected to a multi-stage control switch 2 between these ports.
Connect via 2. The control switch 22 is connected to the next stage control switch 22 via a waveguide 24. The propagation path is
By connecting the outputs of the control switches 22 one after another to the inputs of the control switches 22 in the next stage, it is possible to branch into a tree shape, and as a result, one input port of the tree-type optical switch 20 can connect N optical switches. A propagation path to the output port of is formed. One input port of the first-stage control switch 22 is used to configure the input port of the tree-type optical switch 20, and N output ports of the plurality of final-stage control switches 22 are used to configure the tree-type optical switch. It is sufficient to configure 20 output ports.

It is preferable to use a 1×2 optical switch as the control switch 22, and in this case, one control switch 221Fr of the next stage should be connected to each of the two outputs of the control switch 22 (thereby, the branches of the propagation path All you have to do is form this.

The rear switch 14 is a board +8b, for example, 2 capacitors.
It comprises an NbO3 substrate and an arbitrary suitable number (N) of Nx1 tree-type optical switches 26 provided on this substrate +8b.

The tree-type optical switch 26 has a tree-shaped propagation path that gradually merges from N input ports to one output port. In this embodiment, a plurality of stages of control switches 28 are provided between the input and output ports of the tree-type optical switch 26 to form this propagation path. In each tier,
An arbitrary number of control switches 28 are provided. Then, the N input ports and one output port of the tree-type optical switch 26 are connected through two or more stages of control switches provided between these ports. control switch 2
The control switch 28 is connected to the next stage control switch 28 via a waveguide 24. The propagation paths are merged into a tree shape by successively connecting the outputs of the control switches 28 to the inputs of the control switches 28 at the next stage, and as a result, one A plurality of 0M 1-stage control switches 2 in which a propagation path to the output ports is formed.
If N input ports of 8 are used to configure the input port of the tree type optical switch 26, and one output port of the final stage control switch 28 is used to configure the output port of the tree type optical switch 26. good.

It is preferable to use a 2×1 optical switch as the control switch 28. In this case, one control switch in the previous stage is connected to each of the two inputs of the control switch 28, that is, the two inputs in the previous stage are control switch 2
8 to one control optical switch 28 at the next stage to form a confluence of propagation paths.

The figure shows a configuration example in which N=4 and a 1×2 optical switch is used as the control switch 22 and a 2xl optical switch is used as the control switch 28.

The tree-type optical switch 20 in this case is configured as follows, for example. Control switches 22 are provided in two stages, one control switch 22 is provided in the first stage and two control switches 22 are provided in the second stage. One second-stage control switch 22 is connected to each of the two outputs of the first-stage control switch 22, thereby forming a 1×4 tree-type optical switch 201fr. The front-stage switch 12 has a configuration including four tree-type optical switches 20.

Further, the tree-type optical switch 26 is configured as follows, for example. Control switches 28 are provided in two stages, with two control switches 28 in the first stage and one control switch 28 in the second stage.

One first-stage control switch 28 is connected to each of the two inputs of the second-stage control switch 28, so that four
A ×1 tree-type optical switch 14 is constructed. The latter stage switch 14 has a configuration including four of these tree-type optical switches 26.

For one tree-type optical switch 20, N pieces, for example, 4
The tree-type optical switches 26 are connected without duplication. The output of the tree-type optical switch 20 and the input of the tree-type optical switch 26 are connected in a one-to-one correspondence, so that the front-stage switch 12 and the rear-stage switch 14 are connected in series.

The control switches 22 and 28 are in a cross state or a bar state for TM waves where the electro-optic effect is large.
In order to operate in any state of the Bar r) state, the Cross
Alternatively, it is created to operate only in one of the Bar states.

It is known that when two-cut LiNb○3 substrates are used as the substrates 18a and 18b, the electro-optic effect is greatest for light in the 7M mode, which has a polarization direction perpendicular to the substrate surfaces of the substrates 18a and 18b. There is. Therefore, if the propagation paths of one and the other lights are selectively changed when they are in the 7M mode, the operating voltage can be lowered.

When one light is input to the front switch 12 in the 1M mode, the other light is input in the TE mode with a polarization direction parallel to the substrate surface of the substrate 18a, and the other light is input to the rear switch 14. When inputting light in 1M mode, it is sufficient to input one of the lights in TE mode.

In order to change the polarization direction of the light in this way, a PANDA type polarization maintaining fiber, for example, is used as the polarization changing means 16.

FIGS. 2(A) and (8) are front views showing one end face and the other end face of a polarization maintaining fiber as a polarization changing means, and FIG. One end face of the polarization-maintaining fiber and Figure 2 (
B) shows the other end face of the polarization maintaining fiber connected to the input face of the subsequent switch; arrow 2 indicates substrate 1;
The direction of the Z axis perpendicular to the substrate surface of 8a and 18b is shown.

In these figures, S indicates the axis of the bias changing means 16.

On one end surface 16a and the other end surface 16b, two white circles T1 and T2 are drawn in order to explain the twisted state of the shaft S, respectively, arranged in a line in the diametrical direction. Moreover, when the shaft S is not twisted, the end surface 16a
and +6b When viewed from the direction along the i-axis S, the circle mark T
The arrangement direction of 1 and the arrangement direction of circle mark T2 are drawn in such a way that one defeat occurs.

As mentioned above, in order to change one light from 8TM mode to TE mode and the other light from TE mode to 1M mode, the connected end faces 16a and +6b! When viewed along the axis S, as also shown in FIGS. 2(A) and (B),
They may be connected so that the axis S is twisted by 90' so that the arrangement directions of the circle marks T1 and T2 are orthogonal to each other.

(Configuration of optical switch element) FIGS. 3(A) and 3(B) are diagrams showing an example of the configuration of an optical switch element suitable for use in the control optical switch of this embodiment. This is a concrete example of the configuration of the control switch shown in FIG.

In FIG. 3(A), reference numeral 30 indicates an optical switch element of an inverted Δβ operation type, and this optical switch element 30 includes the following:
The input/output waveguides 32 and 34 are provided on the substrate 18a or +8b so as to be close to each other and parallel to each other in the coupling region 36. The waveguides 32 and 34 are formed by diffusing Ti, for example. Roughly speaking, the waveguide 3 of the optical coupling region 36
The control electrode 44 is divided into any suitable number of electrodes, for example, two.
, and the waveguide 34 of the coupling region 36 is provided with an arbitrary suitable number of control electrodes 46, for example, divided into two. The control electrode 44 is composed of two electrode members 44a and 44b, and the control electrode 46 is composed of two electrode members 46a and 46b.

The electrode members 44a and 44b extend from the right side edge of the waveguide 32 to the left side edge, and the portion extending toward the left side edge extends to the outer region of the waveguide 32.

The electrode members 46a and 46b extend from the left side edge of the waveguide 34 to the right side edge, and further extend the portion extending toward the right side edge to the outer region of the waveguide 34. let

The upper edges of the electrode members 44a and 46a form the coupling region 36.
The upper edge X and the lower edges of the electrode members 44b and 46b meet the lower edge Y of the coupling region 36.

Roughly, the lower right end of the electrode member 44a and the upper left end of the electrode member 46b are electrically connected via the connecting portion 48.

Note that numerals 32a and 32b in the figure indicate the upper and lower ends of the input/output waveguide 32, and numerals 34a and 34b indicate the upper and lower ends of the input/output waveguide 34.

When using the optical switch element 30 configured as described above as the control switch 22, for example, the upper end 32a @
The lower ends 32a and 34b may be used as input ports and output ports. When used as the control switch 28, for example, the upper ends 32a and 34a may be used as % input ports, and the lower end 3
2b% output port may be used.

The input ports and output ports may be set in any suitable manner, with one of the upper ends 32a and 34a being an input port and both lower ends 32b and 34b being output ports,
Alternatively, both the upper ends 32a, 34a can be used as output ports, and either the lower ends 32b, 34b can be used as an input port.

The optical switch element 30 is a cross (Cr) for TM waves.
The optical switch element 30 operates in either the oss or bar state, and is configured so that the optical switch element 30 functions equivalently to a passive branch for TE waves. By configuring the optical switch element 30 to function as a passive branch,
The TE waves input to the
or from the lower end 32b134b).

The optical switch element 30 is driven in the reverse direction Δβ via the control electrodes 44 and 46.

The conditions for creating the optical switch element 30 that operates as described above can be easily set based on the (L/β.)-(ΔβL/rt) diagram. Here, shi indicates the element length, 10 indicates the coupling length, Δβ indicates the propagation constant difference, and ■ indicates the circumference ratio.

FIG. 3(B) is a diagram showing another configuration example of an optical switch element suitable for use in this embodiment. Components corresponding to those of the optical switch element 30 shown in FIG. 3(A) are denoted by the same reference numerals 1, and detailed explanation thereof will be omitted.

In the same figure, 42 indicates inverted ΔB operation and uniform (-like Δβ
1 shows a type of optical switch element that can perform the operation. In this optical switch element 42, the waveguide 3 of the optical coupling region i*36
One control electrode 38 which is not divided into two is provided, and an arbitrary suitable number of control electrodes 40 which are divided into two, for example, are provided in the waveguide 34 of the coupling region 36. The control electrode 40 is composed of two electrode members 40a and 40b.

The control electrode 38 extends from the right side edge of the waveguide 32 to the left side edge, and the portion extending toward the left side edge extends to the outer region of the waveguide 32.

The electrode members 40a and 40b extend from the left side edge of the waveguide 34 to the right side edge, and further extend the portion extending toward the right side edge to the outer region of the waveguide 34. let

The upper edges of the control electrode 38 and the electrode member 40a are connected to the upper edge X of the coupling region 36, and the lower edges of the control electrode 38 and the electrode member 40b are connected to the lower edge Y of the coupling region 36. Let me lose once.

When the optical switch element 42 configured as described above is used as the control switch 22, the upper end 32a may be used as an input port, and the lower ends 32a and 34b may be used as output ports, for example. In addition, when used as the control switch 28 in the latter stage, the upper ends 32a and 34a @input ports and the lower end 3
2b! It can be used as an output port.

The optical switch element 42 is made of Cr for TM waves.

It is configured to operate in either the SS or Bar state, and is configured to function equivalently to a passive branch for TE waves, so that the TE waves input to this element 42 are divided into approximately 1:1 ratios. output with a power transfer ratio of

When obtaining the Cross state for TM waves, the optical switching element 42 is driven in an inverted Δβ manner via the control electrodes 44 and 46, and when obtaining the Bar state for TM waves, the optical switching element 42 is driven uniformly via the control electrodes 44 and 46. ΔB short drive possible.

The operating voltage of the optical switch element 42 can be lower than the operating voltage of the optical switch element 30 shown in FIG. 3(A).

The conditions for creating the optical switch element 42 that operates as described above are CL/I1. ) - (ΔβL/TT) can be easily set based on the diagram.

(Example of Operation) FIG. 4 is an explanatory diagram of an example of the operation of this embodiment, and schematically shows the configuration of each component. In the following description, control switch 22.28! The control switches 22 and 28 selectively operate in the Bar or Cross state depending on the operating voltage for TM waves, and the operating voltage for TE waves. It is assumed that the operation is equivalent to passive branching regardless of the situation.

In the figure, the input port of the tree-type optical switch 20 in the previous stage is designated by reference numeral 20a, and the output port is designated by reference numeral 201.20.
2.203.204, and the input port of the subsequent tree-type optical switch 26 is designated by the symbol 261.262.263.2.
64 and an output port are designated by 26a.

When using 2-cut LiNb○3 substrates as the substrates 18a and 18b, a TM having a bias direction in the biaxial direction of the substrate
The refractive index for the wave and the refractive index for the TE wave, which has a polarization direction perpendicular to the two axes of the substrate, have different values. It propagates as an independent light that does not exchange energy.

Further, when using a two-cut LiNb○3 substrate, the ratio of the electro-optic coefficients γ,3 and γ33 is γ,3/γ33L93.
Therefore, the electro-optic effect for TM waves is approximately 3.2-3.3 of the electro-optic effect for one TE wave.
It will be doubled. Therefore, at a low operating voltage where a Cross or Bar state is obtained for the TM wave, the TE wave input to the control switch 22.28 is adjusted to an ideal power transfer ratio (most ideally 1:0). or O:1), it is not possible to output from two output ports, therefore,
When the TE wave passes through the optical switch element 22 or 28, a loss occurs. Regarding the TM wave, it can be output with an ideal power transfer ratio at a low operating voltage, and almost no loss occurs.

Now, as shown in Figure 4, one light X and the other light Y
However, the input port 20a of the tree-type optical switch 20 in the previous stage
Let us consider the case where the signal is input from the input port 202 to the input port 262 of the tree-type optical switch 26 at the subsequent stage, and is then output from the output port 26a.

As shown in the figure, it is assumed that one light is input as a TM wave and the other light is input as a TE wave to the input port 20a.

Then, the propagation path of one of the input lights X is controlled by the control switch 22 of the tree-type optical switch 20 at the previous stage to selectively change the propagation path as desired, and as a result, it is guided to the output port 202. On the other hand, the control switch 22 is
Since it operates in the same way as a passive branch with a power transfer ratio of approximately 1:1 for the E wave, the other input light Y is
When passing through the control switch 22, the light reaches the output port 202 while generating loss light Yet.

The lights X and Y are then input from the output port 202 to the input port 262 of the tree type optical switch 26 at the subsequent stage via the bias changing means 16. Since the bias changing means 16 twists the axis S by 90 degrees, the light X is changed from TM waves to TE waves, and the light Y is changed from TE waves to TM waves, and the input port 26
2 is input.

The control switch 28 of the tree-type optical switch 26 in the latter stage is T.
Since it operates in the same way as a passive branch with a power transfer ratio of 1:1 for the E wave, the input port 262 is used as the TE wave.
When the input light X passes through the control switch 28, it reaches the output port 26a while causing a loss of light X. On the other hand, the propagation path of the other light input as a TM wave to the input port 262 is controlled by the control switch 28 and selectively changed as desired.
lead to a.

Lights X and Y are then output from the output port 26a to the next stage optical circuit element.

By letting the leaked light X+'& escape into the substrate 18b from a branch that is not used as an output port of the control switch 28, it can be prevented from mixing with the lights X and Y output from the output port 26a.

In addition, leakage light Y is output port 201.203.204
The signal is input from the bias changing means 16 to a tree-type optical switch 26 (not shown) at the subsequent stage. Leak light Y.

is changed into a TM wave by the polarization changing means 16, and is input to another tree-type optical switch 26. To describe this from another perspective, in the later stage tree-type optical switch 26 shown in the figure,
This means that the leaked light Y2 changed into a TM wave is input from the front-stage tree-type optical switch 20 (not shown). However, by letting the leaked light Y2 escape to the substrate 18a from a branch that is not used as an output port of the control switch 28, it is possible to prevent the leaked light Y2 from being mixed with the lights X and Y output from the output port 26a.

As shown in the figure, when the lights X and Y are input to the input port 262 and output from the output port 26a, for example, the control switch 28 of the first stage counting from the input boat side
Both are in the Bar state and the second stage control switch 2
By operating 8 in the cross state, it is possible to prevent leakage light x1 and leakage light Y2 from being mixed in.

Although not shown in the figure, the lights X and Y are input to port 2.
61 and output from the output port 26a, for example, all the first stage control switches 28 are in the Cross state and the second stage control switch 28 is in the Cross state.
By operating in the r08S state, leakage light X and Y2 can be prevented from entering.

Similarly, light X and Y! When inputting to the input port 263 and outputting from the output port 26a, for example, the first
By operating all of the control switches 28 in the first stage in the Bar state and the second stage control switch 28 in the Bar state, it is possible to prevent the leakage lights X+ and Y2 from entering.

Similarly, when inputting light to the X and YIv input ports 264 and outputting it from the output port 26a, for example, both the first stage control switches 281Fr are in the Cr0SS state and the second stage control switch 28 is in the Bar state. By operating in this state, the leakage lights X1 and Y2 can be prevented from entering.

(Conditions for creating the optical switch element) Although the manufacturing conditions are not limited, the following describes the conditions for the optical switch element constituting the control optical switch in order to obtain operating characteristics suitable for practical use in the optical matrix switch snow making of this example. I will explain the creation conditions.

In order to obtain operational characteristics suitable for practical use in this example, it is necessary to: ■ prevent the output optical power of each output port from varying (when outputting one light from either output port, the output of one light Similarly, when the other light is output from any output port, the output optical power of the other light is almost equal). In addition to this, it is preferable that the output optical power of one and the other light be approximately equal to each other when the light is output from the same port.
By realizing these (2) and/or (φ), crosstalk can be reduced and output optical power can be made uniform.

For example, the optical switch element is Cr for the TM wave f.

In order to make the TE wave loss when operating in the SS state and the TE wave loss when operating in the Bar state almost equal, the TE wave loss in the tree type optical switch of the optical matrix switch in the front and rear stages is adjusted. By reducing the fluctuation, the above-mentioned (1) and/or (2) can be achieved. Note that crosstalk and/or loss fluctuations for TM waves can be reduced by operating the optical switch element in an inverted Δβ operation or by operating it in a uniform Δβ operation.

Next, Fig. 5 (A) to (E) and Fig. 6 (A) to (C)
! With reference to this, conditions for creating an optical switch element for reducing TE wave loss will be explained.

Figures 5 (A) to (E) show that the control switch 22.28 is in the third position.
6(A) to (C) are diagrams illustrating manufacturing conditions when the optical switch element 30 shown in FIG. 3(A) is used, and FIGS. 42
FIG. 2 is an explanatory diagram of creation conditions when

In these figures, the horizontal axis represents the ratio of the element length to the coupling length (element length/coupling length), and the vertical axis represents the loss with respect to the TE wave. The solid and broken curves show the loss characteristics of the TE wave when the element length/coupling length for TMi is the value shown by the solid arrow.The solid curve shows the optical switch element in the cross state for the TM wave, and the broken line curve shows the loss characteristics when operating in a bar state for TM waves.

When the element length/coupling length for the TE wave is set to the value (optimal value) shown by the broken line arrow, the TE wave loss in the Cross state and the TE wave loss in the Bar state can be made equal.

As can be understood from these figures, even if the element length/coupling length (hereinafter referred to as the ratio TM) for the TM wave is varied with a width, the element length/coupling length (hereinafter referred to as the ratio □ It can be seen that the optimum value of the ratio TI: is approximately constant. In the case of the optical switch device 30 shown in FIGS. , in the case of the optical switch element 42 shown in FIGS. 6(A) to 6(C),
The optimum values of the ratios are approximately 1.5 and approximately 2.5. When creating an optical switch element, it is sufficient to create the optical switch element so that the ratio T becomes the above-mentioned optimum value.

The set value of the ratio differs depending on the creation conditions and creation method, but the ratio TM, for example, is smaller than the ratio A (about 1/(1,2) to 1/(1,3) of the ratio A). ) times the ratio Tε. In an optical switch element that is created by providing a Ti diffusion waveguide on a LiNbO3 substrate to reduce loss in the waveguide, the ratio TM is usually approximately 1/(1) of the ratio Tε. ,2)～
It becomes 1/(1,3) times.

FIG. 7 is a diagram schematically shown for explaining relaxation of the production conditions. In order to ease the manufacturing conditions, the optical switch element 1
The fluctuation width u of the loss for each TE wave may be set to such a range that the output optical power of the TE wave output from each output port fluctuates within a practically acceptable range. By creating the range V of the ratio T6 that gives U, the creation conditions can be relaxed.

In this example, an optical switch element is created with a range V ratio E such that the variation width U is within 0.4 dB, and this optical switch element is used to make a 16 x 16 type optical matrix switch. Suppose we have configured . In this case, the output optical power of the TE wave fluctuates! It can be within 1.6 dB.

In addition, the curve ■ shown in FIGS. 6(A) to (C) is T
The loss characteristic curve of the TE wave is shown when the optical switch element is operated using a second operation point that gives a Bar state to the M wave. As shown in the figure, the curve H near the intersection with the TE wave loss characteristic curve in the Cross state
Since the slope of is smaller, for example, the range v of the ratio, where the fluctuation width U is within 0.4 dB, can be set wider.

Since the second operating point is used, the operating voltage increases, but since the range of the ratio Tε can be made wider than vV, the manufacturing conditions can be more relaxed.

Furthermore, if the interval between the input and output waveguides of the optical switch element (separation path M) is a variable interval that changes gradually in a tapered manner, the ratio will be such that the fluctuation width U is within 0.4 dB. , 5 can be set wider. Since the intervals are variable, the operating voltage increases, but since the range V of the ratio TE can be made wider, the manufacturing conditions can be further relaxed.

Ratio 7. For TM waves, Ba can be made wider than the range of VT.
This is because any value ratio □6 between the first operating point and the second operating point that gives the r state can be used.

FIG. 8 is a plan view schematically showing the structure of the second embodiment of the first invention. Components corresponding to those in the above-described embodiment are designated by the same reference numerals, and detailed explanation thereof will be omitted.

The optical matrix switch device M50 of this second embodiment includes a front-stage switch 52, a rear-stage switch 54, and a bias changing means 1.
The front stage switch 52 is comprised of N 1×N tree-type optical switches 56, and the rear stage switch 54 is comprised of N.
It is composed of N×1 tree-type optical switches 58.

The tree-type optical switch 56 has a tree-shaped optical propagation path that branches from one input port to N output ports. In this embodiment, in order to form this propagation path, a branching means 60 is provided on the input side of the tree-type optical switch 56 so as to constitute a plurality of stages, and a plurality of branching means 60 are provided on the output side of the tree-type optical switch 56. A control optical switch 62 is provided in parallel.

And one input port of the tree-type optical switch 56 and N
These output ports are connected via a plurality of branching means 60 and a plurality of control switches 62 installed between these boats.

The propagation path in the tree-type optical switch 56 is
The output of the branching means 60 is successively connected to the input of the branching means 60 at the next stage, thereby branching into a tree shape, and the control optical switch 62 is connected to the branching means 60 at the final stage. As a result, a propagation path is formed from one input port to N input ports of the tree-type optical switch 56.

One input of the first-stage branching means 60 is connected to the input port of the tree-type optical switch 56, and the output port of the tree-type optical switch 56 is connected using 1N output ports of the control switches 62 arranged in parallel. Configure.

The branching means 60 includes one input branch and a plurality of output branches, and functions so that the optical powers of the respective lights outputted from the plurality of output branches are approximately equal. It functions in this way for both light, one inputted as a TE wave and the other inputted as a TE wave. Preferably, branching means 60! IX
It is preferable to configure it as a two-branching means, and in this case, it is configured so that the optical power ratio of the respective lights output from the two output side branches (output branches) is approximately 1:1. 1×
The 2-branching means may be, for example, a 1×2 passive branch composed of a figure-7 shaped waveguide.By configuring the figure-7 shaped waveguide symmetrically, the two output sides can be separated. The optical power ratio of branching can be approximately 1:1.

60% branching means I
It is sufficient to form branches of the propagation path by connecting them one by one.

The control switch 62 is an ON-FF type human power switch 62.
a and an output switch 62b.

The input switch 62a is in an OFF state when the light propagation path is not controlled using the control switch 62 provided with the switch 62a, and therefore does not receive one of the lights input as a TM wave. Further, when controlling light using the control switch 62, it is turned on and receives one of the lights input as a TM wave. The output switch 62b is a control switch 62 comprising this switch 62b lFr.
This is to change the propagation path of one of the lights input to the.

Preferably, the control switch 56 is configured as a 1×2 optical switch, in which case the input switch 62a
and the output switch 62b are configured as follows.

The input switch 62a is configured, for example, by using the above-described optical switch element 30 or optical switch element 411X1 as an optical switch element. In this case, by setting the optical switch elements 30 and 42 to the Cross state, the ○N state or the ○FF state can be set.
By setting it to 0.427aBar state, an FF state or an ON state can be formed. In addition, the output switch 62b
For example, the optical switch element 30 or the optical switch element 4
2 as a 1×2 optical switch element.

Input switch 62a and output switch 62bl (tTM
It selectively operates in either the Cross or Bar state for waves. Input switch 62a for TE wave
is equivalent to a single waveguide, and for TE waves, the output switch 62b functions equivalent to a passive branch with an output ratio of approximately 1:1.

The figure shows an example of the structure of a tree-type optical switch 56 in which N=8. In the illustrated example, one 1×2 branching means 60 is provided in the first stage.
, two 1×2 branching means 60 and four control switches 62 are provided in the second stage. One second-stage branching means 60 is connected to each of the two outputs of the first-stage branching means 60. Then, the output of the second stage branching means 60 is
2 control switch 62 is connected. As a result, a 1×8 tree type optical switch 56 is obtained.

Next, the N×1 tree-type optical switch 58 of this embodiment will be explained.

The tree-type optical switch 58 has a tree-shaped light propagation path that gradually merges from N input ports to one output port. In this embodiment, in order to form this propagation path, the control optical switch 64 is replaced with the tree-type optical switch 5.
A plurality of units are provided in parallel on the input side of the day, and a merging means 66
are provided on the output side of the tree-type optical switch 58 so as to form a plurality of stages. The N manual boats and one output port of the tree-type optical switch 58 are connected via a plurality of stages of merging means 66 and a plurality of control switches 64 provided between these boats.

The input of the first-stage merging means 66 is connected to the input of the control light switch 64, and the input of the next-stage merging means 66 is connected to the output of the first-stage merging means 66.The output of the 0-merging means 66 is are successively connected to the inputs of the next-stage merging output 66, and as a result, the propagation paths are merged into a tree shape, resulting in the propagation from the N input ports of the tree-type optical switch 58 to one output port. A path is formed. The input ports of the tree type optical switch 58 are configured using tN input ports of the plurality of parallel control switches 64, and one output port of the final stage merging means 66 is used as the output port of the tree type optical switch 58. Connect to.

The control switch 64 is ON-OFF with the input switch 64a.
It is composed of a type output switch 64b.

The input switch 64a is for changing the propagation path of the other light that is changed into a TM wave and input into the control switch 64 provided with the switch 64a*. The output switch 64b is OF when the control switch 64 including this switch 64b is not used to control the light propagation path.
It enters the F state, and therefore does not output the other input light that has been changed into a TM wave. Further, when the control switch 62 is used to control the propagation path of light, it is turned on, and therefore the other input light is changed into a TM wave and output.

Preferably, it is configured as a control switch 6482 x 1 optical switch, and in this case, the input switch 64a and the output switch 64b are configured as follows.

The input switch 64a is configured by using, for example, the optical switch element 30 or the optical switch element 42M2x1 optical switch element. Further, the output switch 64b is configured by using, for example, the optical switch element 30 or the optical switch element 411.times.1 optical switch element.

The input switch 64a and the output switch 64b selectively operate in either the CroSS or Bar state for the TM wave, and the input switch 64a operates in the CroSS or Bar state for the TE wave.
The output switch 64b functions equivalently to a real waveguide, and equivalently to a merging means.

The merging means 66 comprises a plurality of input branches and one output branch. Preferably, the merging means 66 is constructed as a 2×1 merging means, and in this case, it may be constructed using, for example, a symmetrical Y-shaped waveguide.

The figure shows an example of the structure of a tree-type optical switch 58 in which N=8. In the illustrated example, there are four 2×1 control switches 64 on the input side of the tree-type optical switch 58, and two 2×1 control switches 64 on the first stage.
1 merging means 66 and one 2×1 merging means 66 in the second stage.
I will set it up. One merging means 66 at the first stage is connected to the two control switches 64, and the two merging means 6 at the first stage are connected to each other.
6 is connected to one merging means 66 of the second stage. As a result, an 8×1 tree type optical switch 58 is obtained.

N pieces, for example, eight tree type optical switches provided on the substrate 18a.
The connection between the switch 56 and N (for example, eight) tree-type optical switches 58 provided on the substrate +8b is performed as follows. That is, one tree-type optical switch 56 is connected to N tree-type optical switches 58.
The type optical switches 58 are connected without duplication. A tree type optical switch 58 is connected to each tree type optical switch 56 without duplication, so that the tree type optical switch 58 is connected to each tree type optical switch 56 without duplication.
The output ports of 6 and the input ports of the tree-type optical switch 58 are connected one-to-one.

Is this an optical matrix switch installed in this example? When operating It50, crosstalk can be reduced, for example, as follows. In the following explanation, the propagation path from the input port of the snow making 50 to the output port of the snowmaking 50 will be expressed as a propagation path, and the propagation path other than this propagation path will be referred to as a propagation path M. express.

In operation of the device tf50, only the ON-OFF optical switch 62a of the control switch 62 on the propagation path is turned ON, and the ON-OFF optical switch 62a not on this propagation path! Set to OFF state. By this, T
One light input as an M wave passes only through the control switch 62 on the propagation path.

One of the lights is changed into a TE wave and input to the control switch 64 on the propagation path, but at this time, the tree-type optical switch
Since the output ports of the channels 56 and 58 are connected one-to-one, the leakage light of one light will be transmitted along the other propagation path.
Sutra lol! The other light, which is not mixed into the JM and is input as a TE wave, is input from the output port of the tree-type optical switch 56 to all N tree-type optical switches 58. However, the other light is changed into a TM wave and input to the tree type optical switch 58. Therefore, only the 0N-OFF optical switch 64a of the control switch 62 on the propagation path is set to the ○N state, and the 0N-OFF optical switch 64a that is not on this propagation path
By turning the F optical switch 64a into the OFF state,
The other light can be made to pass only through the control switch 64 on the propagation path. Therefore, it is possible to output one light and the other light only from the output port to be output, and low crosstalk can be achieved.

Furthermore, in this embodiment, the tree-type optical switch 56 is constructed with a branching means 60 and a control switch 62, and the tree-type optical switch 56 is constructed with a control switch 64 and a multiplexing means 66. Therefore, the number of stages of optical switches constituting the tree-type optical switch can be reduced compared to the first embodiment, and therefore the manufacturing conditions can be made more relaxed than the first embodiment (especially ratio, the range of 6 can be made wider).

Second Invention The configuration of the first embodiment of the second invention will be schematically explained below with reference to FIG. 1.

The optical matrix switch device 10 of this embodiment includes a front-stage switch 12 constructed using a compound semiconductor crystal substrate 18a, a rear-stage switch 14 constructed using a compound semiconductor substrate +8b, and these switches 12, +41Fr.
and bias changing means 16 connected in series. For example, the compound semiconductor crystal substrates 18a and 18b are made of GaAs.
Use a substrate.

The front stage switch 12 is composed of N 1×N tree-type optical switches 2.
0 and the subsequent switch 14 has N×1 trees.
It consists of a type optical switch 26.

The tree-type optical switch 20 has multiple stages of control switches 22 provided around the input and output ports of this switch 2o, and similarly the tree-type optical switch 26 has multiple stages of control switches 22 provided between the input and output ports. A control switch 28 is provided. As the waveguide 24, for example, a Rich (Ridc+e) type waveguide is used.

The substrates 18a and 18b are compound semiconductor crystal substrates such as G
It is known that when an aAs substrate is used, the electro-optic effect for TM waves is small and the electro-optic effect for TE waves is large. Here, "the electro-optic effect is small" includes the case where there is substantially no electro-optic effect.

Therefore, one light is input to the front switch 12 in the 1TE mode and the other light is input in the 7M mode.

The propagation path of one of the lights can be selectively controlled by the control switch 22. One of the lights is polarized by the polarization changing means 16.
in the 7M mode through the polarization changing means 16, and the other light
The signal is input to the subsequent switch 14 in TE mode via the TE mode. Therefore, the propagation path of the other light can be selectively controlled by the control switch 28. Therefore, when constructed using a compound semiconductor substrate, switching operation without polarization dependence can be performed using the electro-optic effect.

In this case, the control switches 22 and 28 are configured to operate in either the Cross state or the Bar state for TE waves where the electro-optic effect is large, and to operate in either the Cross state or the Bar state for TM waves where the electro-optic effect is small. For example, in Figure 3 (A
) and the optical switch operation 30.42 shown in FIG. 3(B) can be used to configure the control switch 22.28. The conditions for producing these optical switch elements 30 and 42 may be changed as desired depending on the substrate material.

The light propagation path of the device 10 of the first embodiment is formed in the same manner as the first embodiment of the first invention.

Hereinafter, with reference to FIG. 8, the configuration of the second embodiment of the second invention will be schematically explained.

The optical matrix switch snowmaking 50 of this embodiment includes a front-stage switch 52 constructed using a compound semiconductor crystal substrate 18a IFr, a rear-stage switch 54 constructed using a compound semiconductor crystal substrate 18a, and bias changing means 16.

The front stage switch 52 includes N 1×N tree-type optical switches 5.
618 and the subsequent stage switch 54 are constituted by N N×1 tree-type optical switches 58.

The tree-type optical switch 56 includes a multi-stage branching means 60 and a plurality of control switches 62 provided between the input and output ports of the switch 56.

The branching means 60 is arranged so that the optical powers of the respective lights outputted from the plurality of output branches are approximately equal for both the light f inputted as a TE wave and the other light inputted as a TM wave. It functions as it should.

The control switch 62 is an 0N-OFF type input switch 62
a and an output switch 62b.

The input switch 62a is in an OFF state when the light propagation path is not controlled using the control switch 62 provided with the switch 62a, so that one of the lights input as a TE wave is not input, and the control switch 62 is used to control the light propagation path. When performing control, it is turned on and receives one of the lights input as a TE wave.

The output switch 62b is for changing the propagation path of one of the lights input in the TE wave mode.

Input switch 62a and output switch 62bl (tTE
For waves, the Cross command selectively operates in either the Bar state, for TM waves, the input switch 62a is equivalent to one waveguide, and for TM waves, the output switch 62b is the output It functions equivalently to passive branching with a ratio of approximately 1:1.

Next, the N×1 tree-type optical switch 58 of this embodiment will be explained.

The tree-type optical switch 58 includes a plurality of stages of merging means 66 and a plurality of control switches 64 provided between the input and output ports of the switch 58.

The control switch 64 is ON-OFF with the input switch 64a.
It is composed of a type output switch 64b.

The input switch 64a is for changing the propagation path of the other light that is changed into a TE wave and input into the control switch 64 including the switch 64a. The output switch 64b is this switch 64b? 0FFt when the light propagation path is not controlled using the control switch 64 provided.
When the control switch 62 is used to control the propagation path of the light, it becomes the ON state, and the other light that has been changed to the TE wave and has been input is not output. outputs light.

The input switch 64a and the output switch 64b selectively operate in either the Cross or Bar state for TE waves, and the input switch 64a operates in the 1 state for TM waves.
The output switch 64b functions equivalently to a real waveguide, and equivalently to a merging means.

In this embodiment, the electro-optic effect is utilized by inputting one light into the front switch 12 in TE mode and inputting the other light into the edge switch 14 after changing the TE mode by one. switch operation without polarization dependence.

The light propagation path of the device 50 of this second embodiment is formed in the same way as the second embodiment of the first invention.

The first and second inventions are not limited to the above-mentioned embodiments, and the forming material, forming method, structure, arrangement position, shape, numerical conditions, etc. of each component can be changed as desired.

For example, the substrate material has an electro-optic effect, e.g. LiN
b○, other ferroelectric crystals, GaAS, and other semiconductor materials may be used.

In addition, various tree-type optical matrix switches that have been proposed in the past may be used for the structure of the front-stage and rear-stage optical matrix switches.

Furthermore, the control optical switch may be composed of one optical switch element or a plurality of optical switch elements. Bar state or Cros
It may be configured to operate only in one of the s states, or may be configured to function as a passive branch, or may be configured to function as a 3 dB coupler. When a compound semiconductor substrate is used, it is configured to operate in either the Cross or Bar state for light with a large electro-optic effect, and a 3 dB coupler is used for light with a small electro-optic effect. By configuring it to function as a function, the design becomes easier.

The configuration of the optical switch that makes up the control optical switch is shown in Figure 3 (
The present invention is not limited to the optical switches shown in A) and (B), and various conventionally proposed optical switches may be used, such as a total internal reflection type optical switch or a directional coupling type optical switch.

Further, the polarization changing means may have any suitable configuration that can change the polarization direction of the light.

(Effects of the Invention) As is clear from the above description, according to the first invention of this application, one light is polarized in such a way that switching can be performed at a lower operating voltage; Even if the other light is input to the front matrix switch in a state where it has a polarization direction that requires a higher operating voltage for switching, by changing the polarization direction using the polarization changing means, the switching of the other light can be made more efficient. It can be input to the subsequent optical matrix switch in a state with a bias direction that can be performed at a low operating voltage.

Therefore, by controlling the propagation path of one light in the optical matrix switch in the previous stage and controlling the propagation path of the other light in the optical matrix switch in the latter stage, one and the other light can be controlled at a low operating voltage. Can be switched.

Further, according to the second invention of this application, one light is polarized in a direction in which switching can be controlled by the electro-optic effect, and the other light is polarized in a polarized direction in which switching cannot be controlled substantially by the electro-optic effect. Even if the light is input to the matrix switch, by changing the polarization direction using the polarization changing means, the other light can be input to the subsequent optical matrix switch in a polarization direction that can be controlled by switching using the electro-optic effect.

As a result, in the optical matrix switch at the front stage, the propagation path of one light can be switched and controlled by the electro-optic effect, and in the optical matrix switch at the rear stage, the propagation path of the other light can be controlled by the electro-optic effect. . Therefore, even when constructed using a compound semiconductor substrate, polarization-independent switching using the electro-optic effect can be performed. Since it utilizes the electro-optic effect, the optical matrix switch M has a faster operating speed and consumes less current than an optical matrix switch configured using a conventional current injection type optical switch.
We can provide 1Fr.

[Brief explanation of the drawing]

Figure 1 shows the first embodiment of the first invention and the first embodiment of the second invention.
2(A)-(B) are diagrams for explaining the connection of the bias changing means, and FIG. 3(A)-(8) are the configurations of optical switch elements suitable for use as control optical switches. Figure 4 is a diagram illustrating an example of the operation of the first embodiment; Figures 5 (A) to (E) are preferred conditions for producing the optical switch element shown in Figure 3 (A); 6(A) to 6(C) are diagrams illustrating an example of suitable manufacturing conditions for the optical switch device shown in FIG. 3 CB). FIG. A diagram for explaining relaxation of production conditions, Figure 8 is the second embodiment of the first invention and the second embodiment of the second invention
FIG. 10.50... Optical matrix switch snow making 12.52
-... Front-stage optical matrix switch 14.54-- Later-stage optical matrix switch + 6-... Bias change means,
18a, 18b...Substrate 20-IXN tree-
Type optical switch 22.28.62.64--Control optical switch 24--Waveguide 26-N x 1 Tree-type optical switch Patent applicant
Oki Electric Industry Co., Ltd. ■ ○ Loss to TE wave (dB) 5 (Re) CO> Loss to TE wave (dB) ``0 to E wave
(d8)) ζ q≧
Figure 5: Explanatory diagram of the creation conditions of ζ 9 Figure 5: Explanatory diagram of relaxation of the creation conditions Figure 7: Loss to TE wave (d B) ζ C: h) % (Explanation of the second example and Example 2) Figure S

Claims (2)

[Claims] (1) A front-stage optical matrix switch configured using a ferroelectric crystal substrate and into which two lights with polarization directions orthogonal to each other are input; A rear-stage optical matrix switch that outputs two lights having different directions, and these front- and rear-stage optical matrix switches are connected in series, and the polarization direction at the input surface of the rear-stage optical matrix switch is set to the front-stage optical matrix. and a bias changing means for changing the polarization direction in a direction different from the polarization direction on the output surface of the switch, the control optical switch of the front-stage optical matrix switch is a control optical switch that selects one of the propagation paths of light, and the control optical switch of the rear-stage optical matrix switch In the optical matrix switch device, the control optical switch of the optical matrix switch is a control optical switch that selects the propagation path of the other light, and the preceding stage optical matrix switch is connected to N pieces of 1×N trees.
The optical matrix switch at the latter stage is composed of N x 1 tree optical switches.
An optical matrix switch device characterized in that it is configured with a type optical switch. (2) A front-stage optical matrix switch configured using a compound semiconductor crystal substrate and into which two lights having polarization directions perpendicular to each other are input; A rear-stage optical matrix switch that outputs two lights having the same characteristics as the front-stage optical matrix switch is connected in series with these front-stage and rear-stage optical matrix switches, and the direction of polarization at the input surface of the latter-stage optical matrix switch is set to that of the front-stage optical matrix switch. polarization changing means for changing the polarization direction in a direction different from the polarization direction on the output surface, the control optical switch of the front-stage optical matrix switch is a control optical switch that selects a propagation path of one of the lights, and the control optical switch of the front-stage optical matrix switch selects a propagation path of one light, In an optical matrix switch device, the control optical switch of the switch is a control switch that selects the propagation path of the other light, and the preceding stage optical matrix switch is connected to N pieces of 1×N tree.
The optical matrix switch at the latter stage is composed of N x 1 tree optical switches.
An optical matrix switch device characterized in that it is configured with a type optical switch.