JP2003043417A - Variable difference time delay line and primary polarization dispersion correcting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a variable difference time delay line and a primary polarization dispersion correcting device which can be realized with a fewer number of parts and one double refractive crystal without generating a bias group velocity difference due to insertion of the double refractive crystal, is simple in optical axis adjustment and excellent in productivity. SOLUTION: This variable difference time delay line is provided with the double refractive crystal comprising a biaxial optical monocrystal, and is characterized in that the normal light and abnormal light in the directions of optical axes of the double refractive crystal are made to propagate and the difference time delay of the two polarization directions is corrected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a variable differential time delay line for correcting a group delay time difference due to polarization mode dispersion in an optical transmission system, and a variable differential time delay line using the same.
The present invention relates to a secondary polarization dispersion correction device.

[0002]

2. Description of the Related Art Since the structure of an optical fiber for transmission is not perfectly round, birefringence of about 10 ^-7 to 10 ^-5 occurs. Due to this birefringence, degeneration of two polarization states orthogonal to each other is solved,
Polarization mode dispersion with different group velocity for each polarization (Polarizati
on Mode Dispersion (PMD) occurs. Although this difference in group velocity is very small, it is known that the characteristics are adversely affected by high speed modulation or long distance transmission.

At each optical frequency, there exists an orthogonal polarization state in which the optical frequency dependence of the output orthogonal polarization state is minimized. This state is the PSP (Principal State of Polarizati).
on).

Strictly speaking, the first-order PMD is defined by the differential time delay (DGD: Differential Group Delay) between the output PSPs, but the PSP changes due to disturbance such as temperature change, so the PMD value changes with time. . Also, PS
If P has frequency dependence, it has a higher order (for example, 2
The following PMD is known to occur.

The large time delay that occurs between the different polarization components of the optical signal results in a large optical pulse spread when received. This is particularly noticeable in a system that transmits at a bit rate of 10 Gbs or more per transmission wavelength channel. For example, the spread of the pulse due to the group delay difference of about 20 ps is 0.
Bit error rate (PER) that causes closure of about 5 dB
Bring about a decline. The development of such first-order PMD compensation technology is an urgent task for high-speed long-distance transmission.

As a method of processing a signal failure due to PMD of an optical fiber, an electrical equalization method of signal distortion caused by PMD (reference: IEEE. Photonics Technology Lett. Vol.
2, No. 8, 519, 1990) and an electrical equalization method of the differential time delay in the received electric signal (IEEE. Photonics Technology Let
t., Vol.9, No1., 121, 1997). Such an electrical equalization mechanism can only compensate for the relatively small differential time delay DGD. Also, these mechanisms require expensive high-speed circuit devices.

On the other hand, a method of optically correcting the group delay difference has been considered. There is also a method of making incident on a polarization dependent fiber (PMF) and correcting it by utilizing temperature dependence, but it also has a drawback that the device length becomes large and the response is slow.

There is also a method using chirped fiber Bragg grating (C-FBG: Chirped Fiber Grating) (S. Lee et al., OFC TuS3-1 p272, 2000). C-FBG
One of the disadvantages is that the group delay characteristic has a large wavy phenomenon (group delay ripple) depending on the wavelength. This ripple becomes a problem at higher speeds. There is a need for a variable differential time delay line for polarization mode dispersion compensation that is compact, has no group delay ripple, and has no mechanical sliding parts.

A PMD corrector monolithically integrated using a LiNbO ₃ optical waveguide has also been reported.
(R.Noe et al. Electron Lett., Vol.35, No.8,652,199
9). These have advantages such as integration and high-speed response,
There are several problems such as complicated control structure with hundreds of control points and large polarization dependent loss (PDL).

A method is known in which the signal light is separated into two orthogonal variable components, and only one of the variable components has a different path length to correct the differential time delay (Japanese Patent Laid-Open No. HEI-1).
1-196046).

FIG. 8 shows the configuration thereof. The light transmitted through the optical fiber 120 enters the polarization mode dispersion compensator 150, and the compensated light is converted into an electric signal by the receiver 190. The polarization mode dispersion compensator 125
The automatic polarization controller 130, the variable differential time delay line 150, the optical tap 185, and the distortion analyzer 170 are included.

The polarization composer 130 adjusts the two polarization directions of the PSP so that they coincide with the main polarization axes of the variable differential time delay line 150. In this way, the output signal in the polarization changing period is separated into two orthogonal polarization states for the two PSPs of the fiber, and each delay line 150 is used to delay the two polarization states by a variable amount τc. Correct the time delay difference generated in.

Further, since the PSP of the optical fiber fluctuates with time, it is necessary to automatically correct these in accordance with the fluctuation of the PSP. Therefore, a part of the light is monitored by the optical tap 185 and is coupled to the strain analyzer 170. The strain analyzer 170 measures the strain of the photocurrent and generates an electric output converted into a voltage Vf proportional to the strain.

When the delay time difference between the optical fiber 120 and the compensator 125 is τf and τc, respectively, the total delay time difference can be expressed by the following equation (1). τtotal = r · (τf ^ 2 + τc ^ 2 + 2τf · τc · Cos (2θ)) ^1/2 (1) where 2θ is directly controlled by the PSP of the optical fiber 120 and the polarization controller 130. P of compensation device 125
It is the angle between the so-called Stokes vectors with respect to SP.

When the angle 2θ is adjusted to a numerical value of ± π, the so-called high-speed PSP of the so-called fiber 120 and the so-called high-speed PSP of the so-called fiber 120 are respectively called the high-speed PSP of the compensator 125.
It is clear that the difference time delay sum τtotal is the smallest when │τf-τc│ when it matches SP and the low speed PSP.

Thus, the angle θ adjusts the polarization controller 130 via the feedback path 171 so that the voltage Vf has a relative maximum value.

Similarly, the differential time delay τc in the polarization mode dispersion compensator 125 becomes approximately equal to τf, so that the feedback voltage V becomes zero so that τtotal becomes zero.
It can be adjusted according to the value of f. Thus, if the direction and amount of the differential difference time delay generated by the polarization controller 130 and the variable difference time delay line 150 are adjusted by the above method, the distortion level of the optical signal output by the PMD compensator 125 will be the minimum. Become.

The variable differential time delay line 150 used in this Japanese Patent Laid-Open No. 11-196046 uses a polarization beam splitter to spatially separate the signal light into two polarizations and finally to combine them. Axis adjustment is complicated. Further, since the optical path length is mechanically changed, there is a mechanical sliding portion, and there is a problem in high speed and reliability. Therefore, development of a variable differential time delay difference compensator without a mechanical sliding portion is desired. There is.

[0019]

The object of the present invention can be realized with a smaller number of parts and a single birefringent crystal without generating a bias group velocity difference due to the insertion of a birefringent crystal, and the optical axis can be easily adjusted. It is an object of the present invention to provide a highly variable variable differential time delay line and a first-order polarization dispersion compensator.

[0020]

The present invention comprises a birefringent crystal composed of a biaxial optical single crystal, and propagates the birefringent crystal as an ordinary ray and an extraordinary ray in the optical axis direction. It is a variable differential time delay line characterized by correcting time delay.

When the birefringent crystal has a relationship of nc <na <nb where the refractive index of the three-axis polarized light is (na, nb, nc), b
It is preferable to propagate the light at an angle θ at which the group velocities of the ordinary ray and the extraordinary ray match in a plane including the axis and the c-axis.

Further, it is preferable that a temperature control means for controlling the temperature of the birefringent crystal is provided, and the differential time delay is adjusted by temperature control. Here, it is preferable to provide a light reflecting means for reflecting the light once or a plurality of times on the crystal end face, and increase the effective optical path length by repeating the light reflection.

The optical single crystal having birefringence is preferably formed of potassium neobate. Further, the present invention includes a polarization controller unit for controlling the polarization direction of the light passing through the variable differential time delay line and the optical fiber, and a distortion analyzer unit for analyzing the distortion of the optical pulse. Is a first-order polarization dispersion compensator.

This utilizes the temperature dependence of the refractive index of the crystal axis direction polarized light of the birefringent crystal. By disposing the first birefringent crystal, a group velocity difference bias that is basically proportional to the crystal length L is generated. Therefore, it was necessary to arrange a second crystal of exactly the same length L with their crystal axes orthogonal to each other to offset this bias component. With this arrangement, there is no moving part, and it becomes possible to continuously correct the group velocity difference only by the temperature.

According to the invention, the birefringence of one crystal is utilized as the variable differential time delay line. The principle of the present invention will be described below.

First, regarding the group velocity vg, considering a plane wave in a uniform medium having a dispersive refractive index, the propagation time τ (λ) propagating in a medium of length L is expressed by the following equation (2). To be done. Here, ω is the angular frequency of light. τ (λ) = L / vg (2) Group velocity vg and dispersion (dβ / dλ) are expressed by the following equations (3) and (4). Where β is the propagation constant, k is the wave number, n is the refractive index,
λ is the wavelength. vg = 1 / (dβ / dλ) (3) dβ / dk = d (kn) / dk = n-λ · (dn / dλ) (4) From these equations (2) to (4), The time τ (λ) is expressed by the following equation (5). Here, c is the speed of light. τ (λ) = L ・ (dβ / dω) = L ・ (dβ / dk) ・ (dk / dλ) = (L / c) ・ (n ― λ ・ (dn / dλ))… (5) 2 axes Since the birefringent crystal exhibits different refractive indices (na, nb, nc) in the three crystal axis directions (a axis, b axis, c axis), three propagation times can be defined by the polarization direction. At this time, if nc <na <nb, the extraordinary refractive index in the propagation direction in the plane formed by the c-axis and the b-axis
There is a propagation axis where ne matches the refractive index na of a-axis polarized light. The direction in which the ordinary and extraordinary rays have the same refractive index with respect to the normal phase velocity is called the optical axis. n _e (λ) ＝ n _b (λ) n _c (λ) / √ (n _b (λ) ² Sin ² [θ] + nc (λ) ² Cos ² [θ]) ・ ・ (6) τe (λ) = (L / c) ・ (ne― λ ・ (dne / dλ)) ・ ・ (7) τo (λ) = τa (λ) = (L / c) ・ (na ― λ ・ (dna / dλ))・ ・ (8) Normally, the difference in delay time between the two polarizations is the difference between equations (7) and (8) τe (λ) -τa (λ) = (L / c) ・ (ne-na) -λ・ (L / c) ・ (d ne / dλ-d na / dλ) ・ ・ (9) In order to prevent the increase of the new delay time due to the insertion of the birefringent crystal, the formula (9) is basically 0. Is desirable. The first term on the right side of the equation (9) has a particularly large value, which causes a large bias value. In order to cancel this large bias value, it was necessary to insert a second birefringent crystal having the same crystal length by changing the axial direction by 90 °. However, in the case of a biaxial optical crystal, this (9)
It is possible to prevent the delay time difference of the bias expressed by the equation from occurring. This will be explained below. FIG. 1 shows an index ellipsoid of a biaxial crystal. As an example, it is assumed here that the magnitude of the refractive index is nc <na <nb. FIG. 2 shows a cross section of the index ellipsoid in a plane including the c-axis and the b-axis. The refractive index in the orthogonal a-axis direction does not depend on the propagation axis direction and is 5 in FIG.
It becomes a circle and becomes an ordinary light. Maximum refractive index nb (c-axis propagation,
b-axis polarization) and minimum refractive index nc (b-axis propagation, c-axis polarization)
A propagation axis direction 6 that coincides with the intermediate na exists between and. When considering normal continuous light, the azimuth in which the phase velocities of ordinary light and extraordinary light match is called the optical axis 6. When considering the pulse delay time for a high-speed pulse, it is necessary to consider the group velocity instead of the phase velocity. Time τo
When propagating in the angle θ direction where (λ) coincides, the two polarized waves can propagate at the same group velocity regardless of the crystal length. In order to actually calculate τe (λ) from equation (7), ne
It is necessary to obtain dne (λ) / dλ by differentiating (λ) and ne (λ) by the wavelength λ.

[0027]

[Equation 1]

.. (10) Equation (10) can be calculated by substituting it into (9).
Here, nb '[λ] = d nb [λ] / dλ nc' [λ] = d nc [λ] / dλ ··· (11), which can be calculated from the wavelength dispersion of the refractive indices nb and nc.

The temperature change rate of the propagation time is expressed as follows with respect to the a-axis polarization direction. Δτa (λ) / ΔT = (na −λ ・ (dna / dλ) ・ (1 / c) ΔL / ΔT + (L / c) ・ (δna / δT −λ ・ δ (dna / dλ) / δT) = (na−λ ・ (dna / dλ)) (L / c) α + (L / c) (δna / δT−λ ・ δ (dna / dλ) / δT) ・ ・ (13) where α is a line The coefficient of expansion is usually about 10 ^-7, which is smaller than the refractive index temperature change rate δna / δT by two digits or more, so the first term of the equation (12) can be omitted as follows: Δτa (λ) / ΔT = ( L / c) (δna / δT−λδ (dna / dλ) / δT) ··· (13) For extraordinary light polarization, the temperature change rate of the propagation time can be calculated in the same manner: Δτe (λ) / ΔT = (L / c) (δne / δT−λδ (dne / dλ) / δT) (14) The propagation time difference Δτa−Δτe between the a-polarization component and the extraordinary light polarization component when the temperature is changed by ΔT is as follows. Δτa-Δτe = (L / c) [(δna / δT-δne / δT) -Δ (δ (dna / dλ) / δT- (dne / dλ) / ΔT)] ・ ΔT (15) On the other hand, KNbO ₃ refractive index temperature coefficient of the crystal is not only has a relatively large temperature coefficient at 10 ^-5 order of Refractive index temperature coefficient of the b-axis direction surprisingly is a rare crystal having a negative sign (δna / δT~0.2x10-4,
δnb / δT to -0.35x10 ^-4 , δnc / δT to 0.5
5x10 ^-4 ). Therefore, there is a feature that the group delay time difference between the two polarization directions becomes large due to the temperature change.

Here, potassium neobate (KN: KNbO)
₃ ) The refractive index dispersion of the crystal is (eg: Zysset et.al. J.Op
tical Soc.Am.B, vol9, No3, 380 (1992)).

Utilizing the wavelength dependence and temperature dependence coefficients of these refractive indexes, the propagation azimuth (c-axis and b-axis of the c-axis and b-axis) at which the new delay time difference (9) caused by inserting the birefringent crystal becomes 0 is obtained. It was found that θ = 31.0 ° was obtained by calculating the angle θ) between the wavelengths λ = 1.55 μm. It was found that this is shifted by 1 degree or more with respect to the condition θ = 28.8 ° (optical axis, 6 in FIG. 2) that the phase velocities match. (7 in FIG. 2) Therefore, as shown in FIG. 3, θ = 3 from the c-axis.
The width direction is a so that the 1.0 ° direction is the axis of the length L direction.
The crystal is cut into a rectangular parallelepiped so as to coincide with the axial direction. The main axis of the PSP is a due to the polarization controller not shown in the figure.
It is incident on the crystal so that it is perpendicular to the axis and the a axis. A
The polarized light in the direction perpendicular to the axis becomes extraordinary light, and the change in the a-axis direction becomes ordinary light, and propagates in the crystal 10. Since the crystal 10 coincides with the axis where the group velocities of the ordinary ray and the extraordinary ray coincide with each other, the delay time difference of the bias represented by the equation (9) due to the passage through the crystal 10 does not occur. Here, when the temperature is applied to the crystal by the temperature control element 70, the delay time can be controlled by an amount represented by the equation (15).

Next, when propagating in the azimuth of θ = 31.0 degrees, the temperature dependence of the difference τe−τa in the propagation delay time between the extraordinary light and the ordinary light (a-axis polarized light) at λ = 1.55 μm. The formula (1
The result calculated from 5) is shown in FIG. Where crystal length L =
It is 20 cm.

It has been found that the propagation delay time difference τe-τa can be changed from 0 ps to 4 ps by changing the temperature of the KN crystal from room temperature to 180 ° C.

Considering that about 10% of the pulse interval of the transmission rate is necessary for DGD correction, 40 to 100 Gbp
In the s transmission system, it is possible to correct the delay time difference generated in the optical fiber with this correction amount. Further, the accuracy of the crystal propagation axis direction may be covered in the temperature range even if it is deviated from θ by Δθ. In FIG. 7, θ = 30.0
The temperature dependence of the propagation delay time difference τe−τa is shown. The propagation delay time difference τe−τa at room temperature does not become 0, but can be made 0 ps by adding 100 ° C. as the bias temperature.

To increase the correction amount, a method of lengthening the crystal length may be considered, but in reality, a large KN
There is a limit because crystals cannot be obtained. As a method for solving this problem, the optical path length can be increased by folding the crystal several times.

[0036]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 5 is a configuration diagram showing a first embodiment of the present invention. Normally, in order to match the PSP direction with the polarization direction of the variable differential time delay line before the signal light from the optical fiber is incident on the variable differential time delay line, P
It needs to be controlled by C (Polarization Controller), but it is omitted here.

The light from the fiber is passed through the optical circulator 25.
Incident on the input port 20 of. The output light from the output port 21 of the circulator 25 is collimated by the lens 22. 5 cm long KNbO ₃ as the birefringent crystal 11
Θ = 31 ° from c axis of (Potassium Niobate) to b axis
Is cut out so that the azimuth becomes the propagation axis, and the polarization direction of the output light from the optical fiber (wavelength 1.55 μm) is made to coincide with the a-axis and the direction perpendicular to the a-axis. An optical coating is provided on the side surface of the birefringent crystal 11 opposite to the incident surface so as to achieve total reflection for a wavelength of 1.55 μm, and a total reflection film 50 is provided.

The light reflected by the total reflection film 50 propagates in the opposite direction along the optical axis Q and again has the same optical path 1 in the birefringent crystal.
After passing through 0, the light is again incident on the output port 21 of the circulator by the lens 22. The light incident from the output port 21 of the optical circulator 25 is output from the third port 40. The birefringent crystal 11 is a Peltier temperature element 70.
Set on top of. Birefringent crystal 11 from room temperature to 180
Delay time difference from 0 to 2 by changing the temperature to ℃
It is possible to control up to ps.

FIG. 6 is a block diagram showing a second embodiment of the present invention. Here, an example is shown in which a thin film heater is used to reduce the crystal thickness in order to improve response time and reduce power consumption. The overall structure and crystal arrangement are the same as those in the first embodiment.
5, and the temperature heating method is different from that in FIG. The crystal 11 will be described to avoid duplication.

Using an a-axis propagating potassium neobate crystal (1 mmt × 5 mm × 50 mm) having a length of 5 cm, the crystal c-axis was cut out so that the direction of θ = 31 ° in the b-axis was the propagation axis. The polarization direction of the output light from the optical fiber (wavelength 1.55 μm) is made to coincide with the a-axis and the direction perpendicular to the a-axis. Part of the entrance surface 10a and the exit surface 10b of the crystal 0 is partially provided with a thin film coating 50 that causes total reflection at a wavelength of 1.55 μm. The light incident from the incident surface 10a along the optical axis repeats multiple reflections (four total reflections in the figure) and is emitted from the emitting surface 10b.

A chromium thin film heater 15 is formed on the upper surface of the crystal 10 by vapor deposition or the like, and the thin film heater 15 is energized to control the temperature. Thermal response characteristics can be improved by reducing the crystal thickness to 1 mm.
By using the thin film heater 15 instead of the Peltier element shown in FIG. 5, power consumption can be reduced.

In the above description, an example in which one birefringent crystal is used to compensate the delay time difference is explained, but it can be said that the difference time delay correction amount can be increased by further using these in series. There is no end.

[0043]

As described in detail above, according to the present invention, there is no mechanical sliding portion, neither polarization splitting nor polarization multiplexer is required, and the optical axis adjustment is easy and highly reliable with only one crystal. Thus, it is possible to realize a variable differential time delay line and a first-order polarization dispersion compensator with good response.

[Brief description of drawings]

FIG. 1 is a diagram showing a refractive index ellipsoid of a biaxial crystal.

FIG. 2 is a view showing one section of a biaxial crystal index ellipsoid.

FIG. 3 is a diagram illustrating the present invention.

FIG. 4 is a diagram showing temperature dependence of a propagation delay time difference τe−τa.

FIG. 5 is a diagram showing an embodiment of the present invention.

FIG. 6 is a diagram showing another embodiment of the present invention.

FIG. 7 is a diagram showing temperature dependence of a propagation delay time difference.

FIG. 8 is a diagram showing a conventional example.

[Explanation of symbols]

4 ・ ・ Extraordinary light refractive index, 5 ・ ・ Ordinary light refractive index 6 ・ ・ Optical axis, 7 ・ ・ Directions where the group velocities of extraordinary light and ordinary light match 10, 11 ・ ・ Birefringent crystals 10a, 11a ・ ・ Injection surface 10b , 11b ・ ・ Emitting surface 15 ・ ・ Thin film heater, 20 ・ ・ Input port 21 ・ ・ Output port, 22 ・ ・ Lens 25 ・ ・ Optical circulator 40 ・ ・ Third port, 50 ・ ・ Reflective film 60, 70 ・ ・Temperature control element

Claims (6)

[Claims] 1. A birefringent crystal composed of a biaxial optical single crystal, wherein the birefringent crystal is propagated as ordinary light and extraordinary light in the optical axis direction of the birefringent crystal to correct the difference time delay between two polarization directions. Variable differential time delay line. 2. When the birefringent crystal has a relationship of nc <na <nb where the refractive index of polarized light in three axes is (na, nb, nc), b
2. The variable differential time delay line according to claim 1, wherein the variable differential time delay line is propagated at an angle θ at which the group velocities of the ordinary ray and the extraordinary ray coincide with each other in a plane including the axis and the c-axis. 3. A variable differential time delay line according to claim 1, further comprising temperature control means for controlling the temperature of said birefringent crystal, wherein the differential time delay is adjusted by temperature control. 4. The light reflecting means for reflecting light once or a plurality of times at the crystal end face, and the effective optical path length is lengthened by repeating light reflection. Variable differential time delay line. 5. The optical single crystal having birefringence is formed of potassium neobate.
A variable differential time delay line according to any one of 1. 6. A polarization controller section for controlling the polarization direction of light passing through an optical fiber, and a distortion analyzer section for analyzing the distortion of an optical pulse, according to any one of claims 1 to 5. And a variable differential time delay line of 1.