JP2002082321A - Variable optical filter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize light-speed operation and flexible variable light equalization operation setting the wavelength characteristic of loss in a characteristic having an optional shape. SOLUTION: The variable optical filter is provided with a polarized light separator 6 for polarizing and separating light made incident from an input terminal, two semiconductor optical amplifiers 11a, 11b with gain and the wavelength dependency of the gain changed by injection currents and temperature, current injection parts 12a, 12b and temperature control parts 13a, 13b for individually controlling the gain of the amplifiers 11a, 11b and the wavelength dependency of the gain, by controlling injection currents to the amplifiers 11a, 11b and the operation temperature of the amplifiers 11a, 11b, and a polarized wave multiplexer 10 for polarizing and multiplexing respective light components, outputted from the amplifiers 11a, 11b and outputting the multiplexed light from an output terminal 2.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a variable optical filter for dynamically flattening the wavelength dependence of gain in an optical amplifier.

[0002]

2. Description of the Related Art In recent years, with the practical use of an optical amplifier using an optical fiber doped with erbium as an amplifying medium, wavelength division multiplexing (W) for increasing the transmission capacity of one optical fiber has been proposed.
DM) Transmission systems are being actively developed. In this WDM transmission system, a plurality of optical carriers having different wavelengths are individually modulated, and all the optical carriers are multiplexed by an optical multiplexer. The multiplexed WDM signal is collectively amplified by an optical amplifier, and It is sent to the transmission path. At the receiving end of the optical transmission line, the received WDM signal is separated into individual optical carriers by an optical demultiplexer, and reproduced as data by the optical receiver.

In general, the gain of an optical amplifier has wavelength dependence, and the transmission distance is limited by a signal level deviation between optical carriers caused by the gain deviation. Therefore, in a WDM transmission system, an optical amplifier is cascaded. , It is necessary to minimize the gain deviation in the applicable wavelength band. In order to reduce the gain deviation, a gain equalizer having a wavelength dependence of the transmittance opposite to the wavelength dependence of the gain of the optical amplifier is often used.

As this type of gain equalizer, one using a Fabry-Perot etalon (see Japanese Patent Application Laid-Open No. 9-289349), one using a dielectric multilayer filter, or one using a long-period fiber grating Although these gain equalizers are widely known, the wavelength dependence of the transmittance is fixed, and the gain is changed according to a change in conditions such as a change in the input level to the optical amplifier or a change with time in the system components. Since the wavelength dependence of the variable gain varies, various variable gain equalizers have been proposed.

For example, FIG. 8 is a diagram showing a configuration of a conventional variable gain equalizer. This variable gain equalizer is described in the document "Optical Fiber Conference."
95TuP4 "and" Photonics
“Split-B” by “Technologies”
The "eam Fourier Filter" principle is used. In FIG. 8, the first collimating lens 103a
The second collimating lens 103b is disposed between the light input terminal 101 and the light output terminal 102, and converts light between the first collimating lens 103a and the second collimating lens 103b into collimated light. Further, a glass plate 104 is disposed between the first collimating lens 103a and the second collimating lens 103b, and the movement of the glass plate 104 is controlled by mechanical means 105.

The light incident from the input terminal 101 is
The collimating lens 103a expands the beam diameter of the light and converts it into collimated light. Glass plate 10 inserted in a part of the collimated light beam between first collimating lens 103a and second collimating lens 103b
4, the collimated light has a portion where the glass plate 104 is inserted and a portion where the glass plate 104 is not inserted, and a phase difference occurs between these lights. As a result, the glass plate 104
Interference occurs between light that has passed through and light that has not passed through, and as shown in FIG. 9, it is possible to obtain a wavelength characteristic of periodic loss (transmittance). That is, the amplitude of light can be variably controlled by the ratio of the intensity of light that has passed through the glass plate 104 to the intensity of light that has not passed. Also, by combining several variable gain equalizers using glass plates of different thicknesses, it is possible to obtain a wavelength characteristic of transmittance having an arbitrary shape. In this case, the wavelength characteristic of the light transmittance can be dynamically varied by using a motor or the like as the mechanical means 105 for moving the glass plate 104.

FIG. 10 is a diagram showing another configuration of a conventional variable gain equalizer. This variable gain equalizer is described in the document "Optical Fiber Conference."
2000 WF2 ", which uses electrical means. In FIG.
At the subsequent stage of the input terminal 101, there is disposed a polarization separator 106 that polarizes and separates the incident light L100 into two polarized light beams L101 and L102. In the subsequent stage of the polarization separator 106, a first polarization rotation function utilizing the magneto-optical effect is used.
And a second Faraday rotator 108b. Between the polarization separator 106 and the first Faraday rotator 108a, the polarized light L1
An anti-wavelength plate 107a is provided on a path through which 01 passes. Further, between the first Faraday rotator 108a and the second Faraday rotator 108b, a dielectric polarization filter 109 whose loss wavelength characteristic changes depending on the incident polarization state.
Is provided. At the subsequent stage of the second Faraday rotator 108b, there is provided a polarization multiplexer 110 for multiplexing the two orthogonally polarized lights without causing interference, and the second Faraday rotator 1
08b and the polarization multiplexer 110, and the polarized light L1
A half-wave plate 107b is provided on the path through which the light beam 01 passes. Further, an output terminal 102 for outputting the polarization-combined light is provided downstream of the polarization multiplexer 110.

Here, referring to FIGS. 11 and 12, which are schematic diagrams showing changes in the electric field oscillation of light in each section of the conventional variable gain equalizer shown in FIG. 10, FIG. The operation will be described. First, the description will be made assuming that the incident light L100 from the input terminal 101 has the polarization state shown in FIG. The incident light L100 is converted by the polarization separator 106 into polarized light L10, which is a P-polarized component.
1 is polarized and separated into two polarized components having an orthogonal relationship with the polarized light L102 as the S-polarized component (see FIG. 11B).

The polarized light L101 is directly incident on the first Faraday rotator 108a, and the polarized light L102 passes through the half-wave plate 107a, is converted into the same polarization state as the polarized light L101, and is converted into the first polarized light. Faraday rotator 1
08a. That is, the light enters the first Faraday rotator 108a with the polarization state shown in FIG.

The first Faraday rotator 108a converts the polarized lights L101 and L102a into polarized lights L101b and L102b, respectively, which are rotated by the angle θ (see FIG. 12 (d)). Then, the polarized light L101
b and L102b are dielectric polarization filters 109, respectively.
Is incident on. Here, the dielectric polarization filter 109 is
The S-polarized light and the P-polarized light have the wavelength dependence of the loss shown in FIG. 13, and the polarized lights L101b and L102b output from the first Faraday rotator 108a rotate by an angle θ with respect to the P-polarized light. Therefore, the loss of the light component on the long wavelength side can be larger than that of the light component on the short wavelength side.

The polarized light beams L101b and L102b incident on the dielectric polarization filter 109 are directed in the opposite direction to the first Faraday rotator 108a and are transmitted to the second Faraday rotator 108b to which the same magnetic field intensity is applied. Polarized light L that is incident and is rotated in the opposite direction by the angle θ
101c and L102c, and returns to the polarization state before being incident on the first Faraday rotator 108a (FIG. 12).
(E)).

Thereafter, the polarized light L102c is directly incident on the polarization multiplexer 110, and the polarized light L101c passes through the half-wave plate 107b, is converted into a polarization state orthogonal to the polarized light L102c, and is polarized and multiplexed. It is incident on the vessel 110 (see FIG. 12 (f)). The light combined with low loss by the polarization multiplexer 110 is extracted from the output terminal 102. Note that the first Faraday rotator 108a and the second
The Faraday rotator 108b changes the magnetic field strength applied to an optical crystal that causes the Faraday effect, for example, a bismuth-substituted garnet thick film by an electromagnet or the like, and thereby can variably set the Faraday rotation angle.

[0013]

However, in the conventional variable gain equalizer shown in FIG. 8, the wavelength characteristic of the loss is variably controlled by moving the glass plate 104 by mechanical means 105 such as a motor. Therefore, there has been a problem that high-speed operation cannot be realized and reliability is lacking.

In the conventional variable gain equalizer shown in FIG. 10, the tunable loss wavelength characteristic is determined by the dielectric polarization filter 109 whose loss wavelength characteristic changes depending on the polarization state of incident light. . In general, as shown in FIG. 12, when the wavelength characteristic of loss changes from P-polarization to S-polarization depending on the incident polarization state, the polarization filter has a monotonically rising monotonous characteristic in which the loss on the long wavelength side increases. , Or a monotonous characteristic change that rises to the left where the loss on the short wavelength side increases.
Therefore, the conventional variable gain equalizer shown in FIG. 10 has a problem that it is difficult to obtain a wavelength characteristic of an arbitrary loss.

Further, in the conventional variable gain equalizer shown in FIG. 10, high-speed operation is enabled by changing the intensity of the magnetic field applied to the magneto-optical effect medium using an electromagnet. Is an operation speed of about several microseconds,
There has been a problem that a sufficiently high-speed operation cannot be realized. In the variable gain equalizer, it is desirable that an operation speed of several n seconds or less can be secured.

The present invention has been made in view of the above,
It is an object of the present invention to provide a tunable optical filter capable of realizing a high-speed operation and setting a wavelength characteristic of loss to a characteristic having an arbitrary shape to realize a flexible tunable optical equalization operation.

[0017]

In order to achieve the above object, a variable optical filter according to the present invention comprises: a polarization separating means for polarizing and separating light incident from an input terminal; and a gain and a gain depending on an injection current and a temperature. At least two semiconductor optical amplifiers whose wavelength dependences change, and controlling the injection current and the temperature for the at least two semiconductor optical amplifiers to separately determine the gain of each semiconductor optical amplifier and the wavelength dependence of the gain. Control means for controlling, and polarization combining means for polarization-synthesizing each light output from the at least two semiconductor optical amplifiers and outputting from the output terminal.

According to the present invention, the polarization splitting means splits the polarization of the light incident from the input terminal, and inputs the polarized light to each semiconductor optical amplifier. In each semiconductor optical amplifier, the gain and the wavelength dependence of the gain are individually changed by controlling the injection current and the temperature by the control means. The gain of each semiconductor optical amplifier and the wavelength dependence of the gain are superposed and combined by the polarization combining means, and light filtered by the superposed and combined characteristics is output from the output terminal.

A tunable optical filter according to the next invention comprises: a polarization splitting means for splitting the polarization of light incident from an input terminal;
At least two optical crystals doped with a rare earth whose gain and wavelength dependence of the gain are changed by the injection pump light;
Control means for separately controlling the gain and the wavelength dependence of the gain by injecting the injection pump light into the at least two optical crystals, and controlling each light output from the at least two optical crystals. Polarization combining means for combining the polarization and outputting from the output terminal.

According to the present invention, the polarization splitting means splits the polarization of the light incident from the input terminal and enters the optical crystal. The injection pumping light is incident on each optical crystal, and the gain and the wavelength dependence of the gain are individually changed by the excitation of the added rare earth. The gain of each optical crystal and the wavelength dependence of the gain are superimposed and synthesized by the polarization synthesizing means, and light filtered by the superimposed and synthesized characteristics is output from the output terminal.

A tunable optical filter according to the next invention is characterized in that, in the above invention, the rare earth is erbium, praseodymium, or thulium.

According to the present invention, the rare earth element added to the optical crystal is made of erbium, praseodymium, or thulium, thereby enabling a flexible structure.

[0023] A tunable optical filter according to the next invention comprises a polarization splitting means for polarizing and splitting the light incident from the input terminal;
At least two electroabsorption semiconductors whose wavelength dependence of an absorption coefficient changes according to an applied voltage, control means for individually controlling an applied voltage applied to each of the at least two electroabsorption semiconductors, and the at least two electric fields. Polarization combining means for combining the respective lights output from the absorption type semiconductor and outputting the combined light from an output terminal.

According to the present invention, the polarization splitting means splits the polarization of the light incident from the input terminal, and makes the light incident on each electroabsorption semiconductor. The voltage applied to each electroabsorption semiconductor is
The wavelength dependence of the absorption coefficient corresponding to each applied voltage is controlled by the control means. The wavelength dependence of the absorption coefficient of each electroabsorption semiconductor is superposed and synthesized by the polarization synthesizing means, and light filtered by the superposed and synthesized characteristics is output from the output terminal.

A tunable optical filter according to the next invention is characterized in that, in the above invention, the electroabsorption semiconductor uses the Franz-Keldysh effect.

According to the present invention, the electroabsorption semiconductor changes the absorption coefficient of each electroabsorption semiconductor using the Franz-Keldysh effect so that the wavelength dependence of loss is obtained separately. I have.

A tunable optical filter according to the next invention is characterized in that in the above invention, the electroabsorption semiconductor uses a quantum confined Stark effect.

According to the present invention, the electroabsorption semiconductor changes the absorption coefficient of each electroabsorption semiconductor by using the quantum confined Stark effect so that the wavelength dependence of loss is obtained separately. I have.

[0029]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of a tunable optical filter according to the present invention will be described below in detail with reference to the accompanying drawings.

Embodiment 1 FIG. 1 is a block diagram showing a configuration of the variable optical filter according to the first embodiment of the present invention. In FIG. 1, the incident light L
A polarization splitter 6 for polarization-separating 10 into two polarized light beams L11 and L12 is disposed, and a first semiconductor optical amplifier 11a and a second semiconductor optical amplifier 11a having a low reflection coating on the front and back surfaces of the semiconductor laser. A semiconductor optical amplifier 11b is disposed, and polarized lights L11 and L12 are respectively incident thereon. The first semiconductor optical amplifier 11a includes a current injection unit 12a that injects a controlled current into the first semiconductor optical amplifier 11a and a temperature control unit 1 that controls the operating temperature of the first semiconductor optical amplifier 11a.
3a is connected. On the other hand, the second semiconductor optical amplifier 11
The b is connected to a current injection unit 13b for injecting a controlled current into the second semiconductor optical amplifier 11b and a temperature control unit 13b for controlling the operating temperature of the second semiconductor optical amplifier 11b. At the subsequent stage of the first semiconductor optical amplifier 11a and the second semiconductor optical amplifier 11b, there is provided a polarization multiplexer 10 for multiplexing two orthogonally polarized lights without causing interference. Further, an output terminal 102 for outputting the polarization-combined light is provided at a stage subsequent to the polarization multiplexer 10.

Here, the operation of the variable optical filter shown in FIG. 1 will be described. The incident light L10 incident from the input terminal 1 is separated by the polarization separator 6 into two orthogonal linearly polarized light components, and the first semiconductor optical amplifier 11a and the second semiconductor light are respectively converted into polarized light L11 and L12. The light enters the optical amplifier 11b. First semiconductor optical amplifier 1
In the 1a and the second semiconductor optical amplifier 11b, the injection current and the operating temperature are independently controlled by the current injection unit 12a and the temperature control unit 13a, and the current injection unit 12b and the temperature control unit 13b, respectively. . As a result, in the first semiconductor optical amplifier 11a and the second semiconductor optical amplifier 11b, the gain and the wavelength dependence of the gain are independently changed. Then the first
Semiconductor optical amplifier 11a and second semiconductor optical amplifier 1
The polarized light having the gain and the wavelength dependence of the gain respectively changed by 1b enters the polarization multiplexer 10, is multiplexed, and is output from the output terminal 2.

FIG. 2 shows the wavelength versus loss (gain) dependence obtained by the first semiconductor optical amplifier 11a and the second semiconductor optical amplifier 11b and the wavelength versus loss (light) obtained by the polarization multiplexer 10. FIG. In FIG. 2, FIG. 2A shows the first semiconductor optical amplifier 11.
It shows the wavelength-to-loss dependence obtained by controlling the current injection amount and the operating temperature with respect to a, and has a large loss area on the short wavelength side. FIG. 2B shows the wavelength-to-loss dependence obtained by controlling the current injection amount and the operating temperature with respect to the second semiconductor optical amplifier 11b. Have. Since the polarization multiplexer 10 multiplexes the light output from the first semiconductor optical amplifier 11a and the light output from the second semiconductor optical amplifier 11b, the polarization multiplexer 10 has the wavelength-to-loss dependence shown in FIG. become. In other words, the polarization multiplexer 10 multiplexes the wavelength-to-loss dependence obtained separately by the semiconductor optical amplifiers 11a and 11b, thereby obtaining a complicated loss (gain) wavelength dependence. . For example,
As shown by the dotted line in FIG. 2B, when the wavelength dependence of the loss (gain) caused by the second semiconductor optical amplifier 11b is changed, the wavelength dependence of the loss (gain) is changed.
As shown by the dotted line in FIG. 11 (c), the wavelength dependence of the loss (gain) in which the loss (gain) on the long wavelength side has increased can be finally formed.

In the first embodiment, the case where two semiconductor optical amplifiers 11a and 11b are used has been described. However, the present invention is not limited to this, and three or more semiconductor optical amplifiers may be used.
The injection current and the operating temperature may be controlled separately for each, and the polarization multiplexer 10 may combine the wavelengths so as to further obtain the wavelength dependence of the finer loss (gain).

In the first embodiment, the incident light is separated and multiplexed by the polarization splitter 6 and the polarization multiplexer 10. However, the polarization separator 6 and the polarization multiplexing The device 10 may be realized by an optical coupler or the like using the evanescent effect.

According to the first embodiment, the polarization separator 6
The polarized light beams L11 and L12, which have been polarization-separated by the laser light, enter the two semiconductor optical amplifiers 11a and 11b, respectively, and independently control the current injection amount and operating temperature, respectively, so that different loss (gain) wavelength dependences are obtained. Have
The polarized lights amplified by the semiconductor optical amplifiers 11a and 11b are polarization-synthesized, and a complicated wavelength dependency of a loss (gain) due to superposition can be obtained. Therefore, high-speed operation can be realized since direct control is performed by an electric signal.

Embodiment 2 Next, a second embodiment of the present invention will be described. In the first embodiment described above,
In this embodiment, the injection current and the operating temperature of each of the semiconductor optical amplifiers 11a and 11b are independently controlled to change the wavelength dependence of gain or loss into a desired shape. In No. 2, a plurality of rare-earth-doped optical crystals are independently photoexcited, thereby changing the wavelength dependence of loss (gain) to a desired shape.

FIG. 3 is a block diagram showing a configuration of a variable optical filter according to the second embodiment of the present invention. The variable optical filter shown in FIG. 3 replaces the first semiconductor optical amplifier 11a and the second semiconductor optical amplifier 11b shown in FIG.
A first erbium-doped quartz 14a and a second erbium-doped quartz 14b to which erbium is added are arranged. The first erbium-doped quartz 14a is excited by the excitation light emitted from the optical excitation unit 15a,
The erbium-doped quartz 14b is excited by the excitation light emitted from the optical excitation unit 15b, and independent optical excitation is performed. Other configurations are the same as those of the first embodiment, and the same components are denoted by the same reference numerals.

The first erbium-doped quartz 14a and the second erbium-doped quartz 14b have a diameter of 1.48 μm and a thickness of 0.14 μm.
It is photoexcited by a wavelength such as 98 μm or 0.82 μm which coincides with the energy level of erbium atom, and 1.55
Light in the μm band can be amplified. The incident light L10 entered from the input terminal 1 is polarized and separated by the polarization separator 6 into polarized lights L11 and L12 having two orthogonal linearly polarized components. The polarized lights L11 and L12 are incident on the first erbium-doped quartz 14a and the second erbium-doped quartz 14b, respectively, and are independently excited. As a result, each erbium-doped quartz 14a, 14
In the case of b, independent gain and wavelength dependence of gain are exhibited. Further, each erbium-doped quartz 14a, 14b
The gain and the wavelength dependence of the gain obtained are obtained by combining the gain obtained by each of the erbium-doped quartzes 14a and 14b and the gain obtained by superposing the wavelength dependence of the gain. The wavelength dependence of the gain can be obtained.

FIG. 4 shows the first erbium-doped quartz 14a.
FIG. 3 is a diagram showing a wavelength versus gain dependency obtained by a second erbium-doped quartz 14b and a wavelength versus gain dependency of light obtained by the polarization multiplexer 10. In FIG. 4, FIG. 4A shows the wavelength-gain dependency of the first erbium-doped quartz 14a optically pumped by the optical pumping unit 15a, and has a characteristic that the gain is slightly small on the short wavelength side. FIG. 4B shows the wavelength-gain dependency of the second erbium-doped quartz 14b optically pumped by the light pumping unit 15b, and has a characteristic that the gain is slightly small on the long wavelength side. Since the polarization multiplexer 10 multiplexes the light output from the first erbium-doped quartz 15a and the light output from the second erbium-doped quartz 15b, the polarization multiplexer 10 exhibits the wavelength-gain dependency shown in FIG. become. In other words, the polarization multiplexer 10 multiplexes the wavelength-gain dependency obtained by each of the erbium-doped quartzes 14a and 14b, thereby obtaining a complex gain-wavelength dependency. For example, as shown by the dotted line in FIG. 4A, when the photoexcitation of the first erbium-doped quartz 14a is changed, the wavelength dependence of the gain of the first erbium-doped quartz 14a indicates that the gain is large on the short wavelength side. Characteristics. The wavelength dependence of this gain is polarization-synthesized by the polarization multiplexer 10, and FIG.
As shown by the dotted line, the wavelength dependency of the gain in which the gain on the short wavelength side has increased can be finally formed.

In the above-described second embodiment, the case where two erbium-doped quartzes 14a and 14b are used has been described. However, the present invention is not limited to this. Control the polarization multiplexer 10
May be combined to obtain a more fine-grain wavelength dependency of the gain.

In the second embodiment, erbium is used as a rare earth element to be added to the optical crystal.
However, the present invention is not limited to this, and an optical crystal to which other rare earth elements such as praseodymium and thulium are added may be used.
Further, although quartz is used as the optical crystal, another optical crystal such as a fluoride crystal to which rare earth can be added may be used.

According to the second embodiment, the polarization separator 6
The polarized lights L11 and L12, which have been polarization-separated by the light, enter two erbium-doped quartzes 14a and 14b, respectively.
By independently controlling the optical pumping, the wavelength dependence of different gains is provided, and the polarized lights amplified by the erbium-doped quartz 14a, 14b are polarization-synthesized, and the complex gain-dependent wavelength dependence by superposition is obtained. Therefore, it is possible to flexibly and dynamically flatten the wavelength of the incident light, and to perform high-speed operation by directly controlling the pump light.

Embodiment 3 Next, a third embodiment of the present invention will be described. In the second embodiment described above,
Although the plurality of erbium-doped quartzes 14a and 14b are optically pumped independently to change the wavelength dependence of gain to a desired shape, in the third embodiment, a plurality of electroabsorption semiconductors are used. The wavelength dependence of the loss is changed to a desired shape by independently exciting each other.

FIG. 5 is a block diagram showing a configuration of a variable optical filter according to Embodiment 3 of the present invention. The variable optical filter shown in FIG. 5 includes a first electroabsorption semiconductor 16a and a second electroabsorption semiconductor instead of the first erbium-doped quartz 14a and the second erbium-doped quartz 14b shown in FIG. 16b is arranged. An electric field is applied to the first electroabsorption type semiconductor 16a by the electric field application unit 17a, and an electric field is applied to the second electroabsorption type semiconductor 16b by the electric field application unit 17b. Other configurations are described in Embodiment 2.
And the same components are denoted by the same reference numerals.

The incident light L10 incident from the input terminal 1
Is polarized and separated by the polarization separator 6 into polarized lights L11 and L12 having two orthogonal linearly polarized light components. The polarized lights L11 and L12 are respectively incident on the first electroabsorption semiconductor 16a and the second electroabsorption semiconductor 16b, and an electric field is applied independently. as a result,
In each of the electroabsorption semiconductors 16a and 16b, the wavelength dependence of the absorption coefficient is changed independently. Further, the wavelength dependence of the absorption coefficient changed by each of the electroabsorption semiconductors 16a and 16b is obtained by combining the polarization by the polarization multiplexer 10 and obtaining the wavelength of the absorption coefficient obtained by each of the electroabsorption semiconductors 16a and 16b. The wavelength dependence of the absorption coefficient obtained by superimposing the dependence can be obtained.

Here, the electroabsorption semiconductors 16a and 16b
When an electric field is applied to a semiconductor, a Franz-Keldysh effect occurs in which the absorption edge wavelength shifts to a long wavelength type, and is mainly used for an optical modulator. FIG. 6 is a diagram showing the wavelength dependence of the absorption coefficient in the electroabsorption semiconductors 16a and 16b using the electric field strength as a parameter. In FIG. 6, the characteristics indicated by the solid line indicate the characteristics when no electric field is applied, and the characteristics indicated by the dotted line indicate the characteristics when the electric field is applied. That is, the wavelength dependence of the absorption coefficient shows a characteristic that the absorption coefficient sharply decreases at a predetermined wavelength or more, and shows a characteristic that the predetermined wavelength shifts to a longer wavelength side in accordance with an increase in the applied electric field. . Therefore, the wavelength dependence of the absorption coefficient of each of the electroabsorption semiconductors 16a and 16b can be changed by changing the applied amount of the electric field, and each of the electroabsorption semiconductors 16a and 16b to which the electric field is independently applied can be changed.
The wavelength dependence of the absorption coefficient obtained by the polarization multiplexer 1
By performing polarization synthesis using 0, it is possible to obtain a wavelength dependence of a desired absorption coefficient exhibiting a complicated shape.

For example, FIG. 7 shows the wavelength dependence of the loss obtained by the first electroabsorption semiconductor 16a and the second electroabsorption semiconductor 16b, and the wavelength dependence of the loss obtained by the polarization multiplexer 10. FIG. In FIG. 7, FIG. 7A shows a first example obtained by applying an electric field.
Shows the wavelength dependence of the loss of the electroabsorption semiconductor 16a of FIG.
FIG. 7B shows the wavelength dependence of the loss of the second electroabsorption semiconductor 16b obtained by applying an electric field.
As shown in FIG. 7C, the polarization multiplexer 10 multiplexes the light output from the first electroabsorption semiconductor 16a and the light output from the second electroabsorption semiconductor 16b. This shows the wavelength dependence of the loss with the wavelength dependence superimposed.
Therefore, for example, as shown in FIG.
7 (a) and 7 (b), by further applying an electric field to the electroabsorption semiconductor 16a, the loss on the short wavelength side is increased as indicated by the dotted line.
By superimposing the wavelength dependence of the loss shown in
The wavelength dependence of the loss as shown by the dotted line in FIG. 7C can be obtained. That is, the loss on the short wavelength side can be changed to the wavelength dependence of the loss.

In the above-described third embodiment, the case where two electroabsorption semiconductors 16a and 16b are used has been described. However, the present invention is not limited to this. By controlling the application of an electric field and multiplexing the polarization multiplexer 10, the wavelength dependence of the finer loss may be obtained.

In the third embodiment, an electroabsorption semiconductor using the Franz-Keldysh effect is used. However, the present invention is not limited to this, and an electroabsorption semiconductor using the quantum confined Stark effect may be used. You may.

According to the second embodiment, the polarization separator 6
The polarized lights L11 and L12, which have been polarization-separated by the light, enter the two electroabsorption semiconductors 16a and 16b, respectively, and independently control the application of an electric field, thereby giving different loss wavelength dependences. Polarized light whose loss has been changed by the electroabsorption semiconductors 16a and 16b is polarization-synthesized, and a complicated wavelength dependence of loss due to superposition can be obtained, so that the wavelength of the incident light can be flexibly and dynamically flattened. Since it can be performed and direct control is performed by an electric field, high-speed operation can be realized.

[0051]

As described above, according to the present invention, the polarization splitting means separates the polarization of the light incident from the input terminal, and inputs the polarized light to each semiconductor optical amplifier. In each semiconductor optical amplifier, the gain and the wavelength dependence of the gain are individually changed by controlling the injection current and the temperature by the control means. The gain and the wavelength dependence of the gain of each semiconductor optical amplifier are complicated because they are superposed and combined by the polarization combining means, and light filtered by the superposed and combined characteristics is output from the output terminal. It is possible to flexibly obtain a desired gain and a wavelength dependency of the gain including various shapes, and to achieve an effect of realizing a high-speed operation.

According to the next invention, the polarization splitting means splits the polarization of the light incident from the input terminal and makes it incident on each optical crystal. The injection pumping light is incident on each optical crystal, and the gain and the wavelength dependence of the gain are individually changed by the excitation of the added rare earth. The gain of each optical crystal and the wavelength dependence of the gain are complicated because they are superposed and synthesized by the polarization synthesizing means, and light filtered by the superposed and synthesized characteristics is output from the output terminal. It is possible to flexibly obtain a desired gain and a wavelength dependency of the gain including various shapes, and to achieve an effect of realizing a high-speed operation.

According to the next invention, erbium, praseodymium, or thulium is used as the rare earth element to be added to the optical crystal to enable a flexible structure.
Further, it is possible to flexibly obtain a desired gain including a complicated shape and the wavelength dependency of the gain, and to achieve a high-speed operation.

According to the next invention, the polarization splitting means splits the polarization of the light incident from the input terminal and makes it incident on each electroabsorption semiconductor. The voltage applied to each electroabsorption type semiconductor is individually controlled by the control means, and the wavelength dependence of the absorption coefficient corresponding to each applied voltage is changed. The wavelength dependence of the absorption coefficient of each electroabsorption type semiconductor is complicated because it is superposed and synthesized by the polarization synthesis means, and light filtered by the superposed and synthesized characteristics is output from the output terminal. It is possible to flexibly obtain desired wavelength dependence of loss including various shapes, and to achieve high-speed operation.

According to the next invention, the electroabsorption semiconductor uses the Franz-Keldysh effect to change the absorption coefficient of each electroabsorption semiconductor so as to obtain the wavelength dependence of each loss. Therefore, it is possible to flexibly obtain a desired wavelength dependency of a loss including a complicated shape, and to achieve a high-speed operation.

According to the next invention, the electroabsorption semiconductor changes the absorption coefficient of each electroabsorption semiconductor by using the quantum confined Stark effect so that the wavelength dependence of loss can be obtained separately. Therefore, it is possible to flexibly obtain a desired wavelength dependency of a loss including a complicated shape, and to achieve a high-speed operation.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a variable optical filter according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating the wavelength dependence of loss due to the semiconductor optical amplifier and the polarization multiplexer illustrated in FIG. 1;

FIG. 3 is a block diagram illustrating a configuration of a variable optical filter according to a second embodiment of the present invention;

FIG. 4 is a diagram illustrating wavelength dependence of loss due to the erbium-doped quartz and the polarization multiplexer illustrated in FIG. 3;

FIG. 5 is a block diagram showing a configuration of a variable optical filter according to a third embodiment of the present invention.

FIG. 6 is a diagram showing the wavelength dependence of the absorption coefficient of a general electroabsorption semiconductor.

FIG. 7 is a diagram illustrating wavelength dependence of loss due to the electroabsorption semiconductor and the polarization multiplexer illustrated in FIG. 5;

FIG. 8 is a diagram showing a configuration of a conventional variable gain equalizer.

FIG. 9 is a diagram illustrating the wavelength dependence of loss caused by the variable gain equalizer illustrated in FIG. 8;

FIG. 10 is a diagram showing another configuration of a conventional variable gain equalizer.

11 is an explanatory diagram illustrating a variable gain equalizing operation performed by the variable gain equalizer illustrated in FIG.

12 is an explanatory diagram illustrating a variable gain equalizing operation performed by the variable gain equalizer illustrated in FIG. 10 (part 2).

FIG. 13 is a diagram illustrating the wavelength dependence of the loss of a dielectric polarization filter using the polarization state as a parameter.

[Explanation of symbols]

1 input terminal, 2 output terminal, 6 polarization separator, 10
Polarization multiplexer, 11a first semiconductor optical amplifier, 11b second semiconductor optical amplifier, 12a, 12b current injection unit, 1a
3a, 13b temperature control unit, 14a first erbium-doped quartz, 14b second erbium-doped quartz, 15
a, 15b Photoexcitation section, 16a First electroabsorption semiconductor, 16b Second electroabsorption semiconductor, 17a, 17b
Electric field application unit.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2H049 BA05 BA06 BA08 BB03 BB06 BC25 2H079 AA02 AA13 BA01 CA09 EA07 HA11 5F072 AB09 JJ20 KK15 MM20 YY17 5F073 AB25 AB28 BA01 EA29 GA21 GA24 GA38

Claims (6)

[Claims] 1. A polarization splitting means for splitting polarization of light incident from an input terminal, at least two semiconductor optical amplifiers whose gain and wavelength dependence of gain change according to injection current and temperature, and the at least two semiconductors. Control means for controlling the injection current and the temperature with respect to the optical amplifier to individually control the gain of each semiconductor optical amplifier and the wavelength dependence of the gain; and controlling each light output from the at least two semiconductor optical amplifiers. 1. A variable optical filter, comprising: a polarization combining unit that performs polarization combining and outputs from an output terminal. 2. A polarization separation means for polarization-separating light incident from an input terminal, at least two optical crystals doped with rare earth whose gain and wavelength dependence of gain are changed by injection pump light, Control means for separately controlling the gain and the wavelength dependence of the gain by injecting the injection pumping light into two optical crystals, and polarizing each light output from the at least two optical crystals. And a polarization combining means for combining and outputting from a output terminal. 3. The tunable optical filter according to claim 2, wherein the rare earth is erbium, praseodymium, or thulium. 4. A polarization separation means for polarization-separating light incident from an input terminal, at least two electroabsorption semiconductors whose wavelength dependence of absorption coefficient changes according to an applied voltage, and at least two electroabsorption semiconductors. Control means for individually controlling an applied voltage to be applied to each of the semiconductors; and polarization combining means for polarizing and combining each light output from the at least two electroabsorption semiconductors and outputting the light from an output terminal. Variable optical filter characterized by the above-mentioned. 5. The variable optical filter according to claim 4, wherein the electroabsorption semiconductor uses the Franz-Keldysh effect. 6. The semiconductor device according to claim 4, wherein the electroabsorption semiconductor uses a quantum confined Stark effect.
3. The variable optical filter according to item 1.