JP4550630B2 - Variable dispersion compensator - Google Patents

Description

  The present invention relates to a tunable dispersion compensator used in the field of optical communication.

  In general, an optical transmission line that performs optical communication (optical transmission) has positive chromatic dispersion in the optical transmission band. In order to suppress distortion of signal light due to this chromatic dispersion, conventionally, a dispersion compensating optical fiber having negative chromatic dispersion in the optical transmission band is connected to the optical transmission line to compensate for the chromatic dispersion of the optical transmission line. . This dispersion compensation technique is being put into practical use in optical transmission where the transmission speed of signal light is 10 Gbit / s or less.

  However, the above-described dispersion compensation technique is difficult to apply when the transmission speed of the signal light is further increased. For example, when the transmission speed of signal light is 40 Gbit / s or more, a more accurate dispersion compensation technique is required than the above-described dispersion compensation technique.

  Further, the dispersion compensating optical fiber described above is configured to control the amount of dispersion compensation by the length of the optical fiber. For this reason, dispersion compensation is difficult when the dispersion compensation amount is small, such as several ps / nm to several tens ps / nm, and dispersion compensation corresponding to the above-described high-speed transmission of 40 Gbit / s or more is difficult. In addition, in order to perform dispersion compensation of an optical transmission line using a dispersion compensating optical fiber, it is necessary to form a dispersion compensating optical fiber having a length corresponding to each optical transmission line so as to correspond to each optical transmission line. However, it is not preferable in terms of cost and size.

  Therefore, in order to cope with the situation as described above, a variable dispersion compensator capable of changing the dispersion value with a configuration different from the conventional configuration using the dispersion compensating optical fiber has been studied. As this tunable dispersion compensator, for example, a lattice filter type tunable dispersion compensator formed of a planar optical waveguide circuit (PLC) has been proposed (Non-Patent Document 1).

  An example of a tunable dispersion compensator using a lattice filter is shown in FIG. The lattice filter type dispersion compensator is configured by alternately connecting a symmetric Mach-Zehnder interferometer (MZI) 3 and an asymmetric MZI 4 composed of two optical waveguides 1 and 2, and an interference arm of each MZI 3 and 4. A phase shifter 5 is formed in the part to make the coupling rate and phase variable.

The frequency characteristic of such a lattice filter is obtained by Fourier transform of impulse time response, and is expressed as follows using a Fourier series.

Where N is the number of taps of the optical filter, j = √ (−1), n _{eff is} the equivalent refractive index of the waveguide, f is the optical frequency, ΔL is the optical path length difference of the delay circuit, and g _n is the tap coefficient. is there. In this optical filter, by changing the tap coefficient g _n , that is, the phase shift value of the optical filter, it is possible to realize arbitrary filter characteristics with the upper limit being a value determined by the tap number N. In addition, characteristics appear periodically for each free spectral range (FSR) determined by the optical path length difference ΔL.

  Using such characteristics, a variable dispersion compensator as in Non-Patent Document 1 can be configured using a lattice filter. Lattice filters have the advantage that there is no fundamental loss and loss variation because light is always confined in two waveguides.

  On the other hand, as proposed in Non-Patent Document 1, even when a transversal filter as shown in FIG. 25 is used, filter characteristics similar to those of a lattice filter can be obtained, and a variable dispersion compensator can be configured. .

  As shown in FIG. 25, the transversal filter branches input light into N paths (taps) using variable optical splitters # 1 to #n, adds a delay ΔL between the paths, and further passes each path. The phase shifters ψ1 to ψn + 1 are formed in the optical path, the light intensity and the phase of each path are varied, and the light is combined and output by the optical multiplexer k.

  In this transversal filter, since the Fourier coefficient directly corresponds to the light intensity and phase of each path, it is easy to set a desired characteristic as compared with the Fourier filter, and the number of taps N is increased to increase the resolution of the filter characteristic. Even when the delay circuit is made larger, the delay circuits are arranged in parallel, so that there is an advantage that the delay circuits can be reduced in size as compared with the Fourier filters in which the delay circuits are connected in cascade.

Koichi Higuchi, “Development of planar lightwave circuit to optical functional devices”, Applied Physics Vol. 72, No. 11 (2003), pages 1387 to 1392

  However, when a tunable dispersion compensator is configured using this lattice filter, there is a problem that it is difficult to set desired characteristics because the Fourier coefficient and the coupling rate of each MZI do not correspond directly. Further, when the number of taps N is increased in order to improve the resolution of the filter characteristics, the circuit length increases in proportion to the number of taps N, which makes it difficult to reduce the size.

  On the other hand, the transversal filter has a problem in that loss variation occurs depending on the phase setting condition.

  SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and its object is to provide a variable dispersion compensator that is small in loss and loss fluctuation, can easily set desired characteristics, and can be miniaturized even when resolution is increased. Is to provide.

In order to achieve the above object, the present invention has the following configuration as means for solving the problems. That is, the present invention has an optical waveguide circuit formed on a substrate, and the optical waveguide circuit is a multi-stage optical branching coupler having a tree structure in which an optical branching coupler is arranged at each node, The structure is such that one optical branching coupler corresponding to the root node is the first stage, the light input to the one optical branching coupler is branched and output, and each of the branched and output lights is separated in the next stage. The multi-stage optical branch coupler, which has a structure in which the number of branches increases as the stage progresses, and the light output from the final stage of the multi-stage spectral coupler are respectively input, and the propagation time of the propagation light is calculated. A multi-stage optical multiplexing coupler having a plurality of optical delay lines that are output after being delayed by a set time, and an inverted tree structure in which an optical multiplexing coupler is disposed at a position symmetrical to the optical branching coupler with the optical delay line in between. In the inverted tree structure, a plurality of optical multiplexing couplers each having an optical input terminal connected to the optical delay line in a one-to-one manner are used as the first stage, and light input to the first stage optical multiplexing coupler is multiplexed. Each of the combined and output light is input to a separate optical multiplexing coupler in the next stage, and the number of branches decreases as the stage progresses, and light is output from one optical multiplexing coupler in the final stage. The multi-stage optical multiplexing coupler having the structure and the optical delay line are arranged in parallel with each other at an interval so that the length is increased by a set amount in order from one end side to the other end side in the parallel arrangement direction. An optical transversal filter circuit is configured, and one or more of the optical branching couplers are provided with an optical branching ratio adjusting means capable of changing an optical branching ratio, and one or more of the optical multiplexing couplers can change an optical multiplexing ratio. One optical multiplexing ratio adjustment means is provided. The said optical delay line of the upper is provided a variable capable optical phase adjusting means the phase of the propagating light, each branch of said optical branch coupler and said optical multiplexing coupler are disposed at positions symmetrical across the optical delay line The intensity ratio of the light and the intensity ratio of the combined light are adjusted to be equal to each other , and input from the optical input terminal (optical input port) of the optical branching coupler at the first stage and the optical multiplexing coupler of the final stage. Light that is output from the optical output terminal (optical output port) is adjusted so that the pass wavelength characteristic of the light has a desired group delay characteristic within a predetermined pass wavelength band. Are split into two orthogonal polarization mode lights, output from the first and second polarization ports, and further input from the first and second polarization ports. The common port A waveguide-type polarization splitter / combiner that is output from the first polarization port, a half-wave plate is inserted into the first or second polarization port, and the first polarization port is the first-stage optical branching coupler. The tunable dispersion compensator is connected to the optical input terminal of the optical fiber, and the second polarization port is connected to the optical output terminal of the optical multiplexing coupler at the final stage.

  According to the present invention, it is possible to form a tunable dispersion compensator that is small in loss and loss variation, can easily set desired characteristics, and can be downsized even when the resolution is increased.

  Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 shows a configuration diagram of a tunable dispersion compensator according to the present invention.

  As shown in FIG. 1, the tunable dispersion compensator of the present embodiment is made of a quartz-based planar optical waveguide, and is an optical branching coupler VCx-x (xx is 2) consisting of a 2-input 2-output MZI circuit. Optical multiplexing consisting of a multi-stage optical branching coupler with 8 optical output terminals and a 2-input 2-output MZI circuit, formed by connecting three stages of arbitrary values between -1 and 4-4 in a tree shape A multi-stage optical multiplexing coupler VCy-y (y is an arbitrary value between 5-1 and 8-1) having eight optical input ends, formed by connecting couplers in three stages in an inverted tree; Eight interleaved between the respective optical output terminals of the multistage optical branching coupler VCx-x and the corresponding optical input terminals of the multistage optical multiplexing coupler VCy-y, which delay the propagation time of the propagation light by a set time. An optical delay line d, and the optical delay lines d are arranged in parallel to each other with a gap therebetween and directed from one end side to the other end side in the parallel arrangement direction. The length in the order Te is constituted by an optical transversal filter circuit formed to be longer by a set amount.

  Each of the optical branching couplers VCx-x and VCy-y includes two directional couplers 13 and 14 formed by bringing two optical waveguides 11 and 12 close to each other as shown in FIG. The first and second connecting waveguides 11 ′ and 12 ′ connecting the two directional couplers 13 and 14 are connected to the two connecting waveguides 11 ′ and 12 ′ by a thermo-optic effect. A thin film heater 15 is formed as an optical branching ratio adjusting means for adjusting the branching ratio by changing the refractive index of the waveguide and changing the optical path length difference between the two connecting waveguides 11 ′ and 12 ′.

  Further, on each optical delay line d, there is formed a thin film heater which is a phase adjusting means capable of changing the phase of propagating light propagating through each optical delay line d by the thermo-optic effect, and constitutes phase shifters PS1 to PS8. Yes.

  Each of the thin film heaters is supplied with electric wiring made of a gold thin film, but the illustration is omitted here for the sake of simplicity.

  In the optical transversal filter circuit, one of the two optical input ends of the first-stage optical branching coupler VC2-1 is connected to the chip end surface as the optical input port 104, and the other is connected to the chip end surface as the monitor port 105. It is connected to the. Of the two optical output ends of the first-stage optical multiplexing coupler VC8-1, one corresponding to the optical input port 104 is connected to the chip end face as the optical output port 204, and the other as the monitor port 205. Connected to chip end face

  Of the two optical input terminals in the second and subsequent optical branching couplers VCx-x, the optical input terminals that are not used for connection to the preceding optical branching coupler VCx-x are monitor ports 101 to 103 and 106 to 108, respectively. Connected to the chip end face. Similarly, of the two optical output terminals in the optical couplers VCy-y at the second stage and thereafter, the optical output terminals that are not used for connection with the optical coupler coupler VCy-y at the previous stage are the monitor ports 201 to 203 and 206. ˜208 are connected to the chip end face.

  Further, in order to reduce the power consumption of the heater, heat insulating grooves 16 are formed on both sides of each thin film heater by removing the optical waveguide film until it reaches the substrate.

  With such a configuration, the energization amount to the thin film heater 15 on each MZI circuit constituting the multistage optical branching coupler VCx-x and the multistage optical multiplexing coupler VCy-y is appropriately adjusted, and the coupling rate of each MZI circuit is changed. Thus, the light amplitudes of the eight light paths (taps) can be arbitrarily adjusted. Moreover, the phase of each tap can be arbitrarily adjusted by appropriately adjusting the energization amount to the thin film heater on each optical delay line d.

  The dispersion compensator was manufactured as follows.

  First, an optical transversal filter circuit composed of a silica-based optical waveguide was formed on a silicon substrate by using a flame hydrolysis deposition method (FHD method) and reactive ion etching (RIE). The relative refractive index difference of the waveguide was 1.5%, and the core size was 5 μm × 5 μm. Next, a thin film heater and a power supply electrode were formed by sputtering. Next, a heat insulating groove was formed by RIE. Finally, the chip was cut out by dicing.

Table 1 shows various parameters used to configure this optical transversal filter circuit.

  Next, a method for determining the optical amplitude and phase of each tap for obtaining desired dispersion characteristics will be described below.

  Now, assuming that the number of taps (total number of optical delay lines) of the optical transversal filter is N (N is an integer of 3 or more), the transfer function of the optical transversal filter is expressed by the equation (Equation 2).

Here, α _n and β _n are the ratios of the optical electric field amplitudes of the multistage optical branching coupler and the multistage optical multiplexing coupler, respectively, and the square of the optical electric field amplitude is the light intensity. Φ _n is the phase change amount of the optical phase adjusting means PS1 to PS8, ΔL is the optical path length difference of the optical delay line, n _eff is the equivalent refractive index of the optical waveguide, f is the optical frequency, c is the speed of light, j Is √ (-1).

  Further, n is a tap number, and this number is an array number of the optical delay line. Here, in order from the optical delay line with the shortest length, 0, 1, 2,... Are set, and the number of the optical delay line with the longest length is N-1.

Here, in order to reduce the insertion loss of the optical transversal filter, when α _n = β _n and γ _n = α _n ² = β _n ² , (Equation 2) can be rewritten as Equation (Equation 3).

Assuming that the equations (Equation 4) to (Equation 6) are used (where m is a positive integer), and further the tap coefficient g _n = γ _n exp (jΦ _n ), the equation (Equation 7) is obtained and the frequency Can be discretized in a region.

Here, if the desired frequency characteristic is G _l and the number of samplings is N ′, the tap coefficient g _n shown in the equation (Equation 8) is obtained from the equation (Equation 6) by the following discrete Fourier transform, and the equation ( As in (Equation 9) and (Equation 10), γ _n and the phase change amount Φ _n are obtained.

In this way, it is possible to determine the tap coefficient g _n that satisfies the desired target properties.

  In the present embodiment, since the variable dispersion compensator is configured, the target characteristics include a characteristic that the wavelength characteristic of the loss in the pass band is flat and low, and exhibits a constant wavelength dispersion in the pass band. desirable. Since the chromatic dispersion is given by the second derivative of the phase, a quadratic curve characteristic is desirable in the pass band as the phase characteristic.

Therefore, using the definition equation shown in the equation (Equation 11), in which the loss in the passband is zero and the phase is a quadratic curve with the passband center at the top, the coefficient ε is in the range of −1.5 to 1.5. The target characteristics were defined by changing in.

  Here, λ: wavelength, λc: center wavelength, and ε: coefficient.

  An example of the target characteristics obtained in this way is shown in FIG. The example of FIG. 3 is for the case where the coefficient ε = 0.4.

Next, the tap coefficient at each ε was obtained by the discrete Fourier transform of equation (Equation 8). The results are shown in Table 2. As can be seen from Table 2, the relationship between the array number of the optical delay line and the intensity and phase of the light passing through each optical delay line is symmetric about the N / 2 optical delay line number. When N is an odd number, the optical delay line array number of (N-1) / 2 is symmetric. That is, the tap number = 0 in Table 2 is omitted.

  Subsequently, the loss and group delay spectrum obtained when the number of taps = 8 was calculated using the obtained tap coefficients. FIG. 4 shows the obtained group delay spectrum, FIG. 5 shows the loss spectrum, and FIG. 6 shows a graph showing changes in chromatic dispersion with respect to the coefficient ε.

  FIG. 4 shows that the slope of the group delay spectrum, that is, the chromatic dispersion changes according to the change in the coefficient ε. Further, it can be seen from FIG. 5 that the loss fluctuation within the pass band of about 0.6 nm is suppressed to within 3 dB. Further, as shown in FIG. 6, it can be seen that ± 92 ps / nm is obtained as the value of chromatic dispersion when the coefficient ε = ± 1.5. For both group delay and loss, a similar waveform appears at a period of the FSR wavelength interval (wavelength interval = about 0.8 nm, frequency interval = 100 GHz).

  Next, the actual variable dispersion compensator is controlled using the obtained tap coefficient. In order to accurately set the obtained tap coefficient, light is passed through the monitor port and each optical branching coupler and optical multiplexing coupler can be changed. The branching characteristics were measured.

  For example, when measuring the characteristics of the optical branching coupler VC2-1, LED light having a wavelength of 1.55 μm is input from the monitor port 107 while changing the energization amount to the thin film heater of the optical branching coupler VC2-1. The output light intensity was measured. Since only one of the lights branched by the optical branching coupler VC2-1 is output to the monitor port 207, the light intensity changes according to the variable branching characteristics of the optical branching coupler VC2-1, as shown in FIG. Characteristics can be obtained. If this is converted into a branching ratio, it becomes as shown in FIG. 8, and this can be used to determine the energization amount for setting an arbitrary branching ratio.

  On the other hand, when both of the two output ports of the optical branching coupler are connected to one optical multiplexing coupler, only one of the branched lights cannot be measured. For example, in the case of the optical multiplexing coupler VC4-1 The light intensity input from the monitor port 101 and output to the monitor port 201 is measured. In this case, since the light branched by the optical multiplexing coupler VC4-1 propagates through the phase shifters PS1 and PS2 to the monitor port 201 and is output again after being combined by the optical multiplexing coupler VC5-1. There arises a problem that the interference between the two interferes with the measurement.

  However, this time, the optical path length difference between the two optical delay lines connecting the optical branching coupler and the optical multiplexing coupler is sufficiently longer than the coherent length of the LED light used for measurement. Since it did not occur, it was possible to measure without problems.

  Similarly, the relationship between the energization amount and the branching ratio is measured for other optical branching couplers and optical multiplexing couplers, and the energizing amount for obtaining an arbitrary branching ratio can be obtained for all optical branching couplers and optical multiplexing couplers. I did it.

  Next, in order to grasp the characteristics of each phase shifter, the light of the wavelength tunable light source is input from the monitor port, and the branching ratios of the multistage optical branching coupler and the multistage optical multiplexing coupler are appropriately adjusted, so that only two phase shifters adjacent to each other The interference spectrum was measured while changing the amount of current applied to one phase shifter in a state where the input light was distributed with a light intensity equal to.

  For example, when measurement is performed by energizing the phase shifter PS4 while passing light between the phase shifters PS4 and PS5, the input light from the optical input port 104 is measured and the output light from the port 204 is measured, and the optical multiplexing coupler VC2- 1 and the optical multiplexing coupler VC7-1 at a coupling rate of 50%, the optical multiplexing couplers VC3-1, VC3-2, VC6-1, VC6-2 at a coupling rate of 0%, the optical multiplexing couplers VC4-2, VC4-3 , VC5-2, and VC5-3 are set to a coupling rate of 100%, so that light is branched to the phase shifter PS4 and the phase shifter PS5 at a ratio of 50%.

  The measurement results are shown in FIG. 9, and the relationship between the peak wavelength and the energization amount shown in FIG. 9 is shown in FIG. 9 and 10, it can be seen that a wavelength shift occurs with a change in the energization amount and overlaps with an adjacent peak at the time of no energization due to energization of about 650 mW. That is, the phase shifts by one rotation (2π) by energization of about 650 mW. From this result, the energization amount to the phase shifter PS4 for setting an arbitrary phase shift amount can be obtained.

  Similarly, the characteristics of other phase shifters were also measured.

  However, when comparing the spectrum measurement results between all adjacent phase shifters in the non-energized state, the optical path length difference between adjacent phase shifters is a constant value, so all waveforms should match. However, it was found that the waveforms do not coincide as shown in FIG. This is presumably because the effective optical path length difference slightly varies due to manufacturing process errors.

  Therefore, using the phase shifter driving characteristics measured as described above, each phase shifter was energized so that all the waveforms matched to correct the error. The results are shown in FIG.

  FIG. 12 shows that the waveforms measured between the adjacent phase shifters match. In the subsequent measurement, it was possible to accurately set a desired phase by giving a phase shift with this state as an initial value.

  Using the characteristics obtained as described above, the energization amounts of the respective optical branching couplers, optical multiplexing couplers, and phase shifters were set so as to satisfy the respective tap coefficients in Table 2.

  When setting the branching ratio of each optical branching coupler and optical multiplexing coupler, for example, as in the optical multiplexing coupler VC4-1 and the optical multiplexing coupler VC5-1, the optical branching coupler and the optical multiplexing at symmetrical positions with the optical delay line in between. The coupler branching ratio is set to be equal so that excessive loss does not occur during multiplexing.

  FIG. 13 shows the group delay spectrum measured in this way, FIG. 14 shows the loss spectrum, and FIG. 15 shows a graph showing the change in chromatic dispersion with respect to the coefficient ε.

  From FIG. 13, it was confirmed that the slope of the group delay spectrum, that is, the chromatic dispersion changes according to the change of the coefficient ε. Further, from FIG. 14, the minimum loss in the pass band of about 0.6 nm when the coefficient ε = ± 1.5 is about 10.4 dB, and the loss fluctuation is about 3.5 dB. Moreover, the same waveform could be confirmed with a period of FSR = 0.8 nm (frequency interval = 100 GHz) for both group delay and loss. As a value of chromatic dispersion, as shown in FIG. 15, about ± 90 ps / nm was obtained at a coefficient ε = ± 1.5.

  Note that all the optical characteristics were measured by inputting only TE polarized light into the tunable dispersion compensator using a polarizer and a polarization maintaining fiber. This is because the tunable dispersion compensator of this embodiment example has polarization dependency.

  Therefore, when the tunable dispersion compensator of this embodiment is applied to a system in which arbitrary polarization propagates, propagation light is inserted into the polarization splitter / combiner 22 through the circulator 21 as shown in FIG. And TM polarization light, and then using the polarization maintaining fiber 23, TE polarization light is input from the optical input end of the tunable dispersion compensator 24, and TM polarization light is converted into TE polarization light and then tunable dispersion compensation is performed. This is used by using so-called polarization diversity technology in which each polarized light that has been input from the optical output terminal of the optical device 24 and passes through the tunable dispersion compensator 24 is again combined by the polarization splitter / combiner 22 and output through the circulator 21. Just do it.

  Next, a second embodiment of the present invention will be described with reference to the drawings. FIG. 17 shows a configuration diagram of a tunable dispersion compensator according to the present invention.

  The configuration of the tunable dispersion compensator according to the present embodiment is almost the same as that of the tunable dispersion compensator according to the first embodiment, but the four-stage optical branching coupler VCx-x (xx is 1-1 to 4-8). 16-tap optical transversal filter circuit composed of an optical multiplexing coupler VCy-y (y is an arbitrary value between 5-1 and 8-1) and 16 phase shifters PS1 to PS16 It consists of

  In the tunable dispersion compensator according to the present embodiment, a waveguide polarization splitter / combiner 31 and a half-wave plate 32 are integrated on a quartz optical waveguide chip constituting the tunable dispersion compensator.

  Thus, by increasing the number of taps to 16, the resolution of the transversal filter is increased, and a larger wavelength dispersion can be obtained, and the waveguide-type polarization splitter / combiner 31 and the half-wave plate 32 are integrated. As a result, the number of parts can be reduced and the size can be reduced. However, since the number of intersections with the monitor waveguide increases as shown by the broken lines in FIG. 17, an increase in radiation loss at the intersections is a problem. It becomes.

  As can be seen from FIG. 17, the cross waveguides exist in the optical transversal filter circuit of this embodiment because the connection waveguides or polarizations connecting the optical branching couplers VCx-x or optical multiplexing couplers VCy-y. This is the intersection of the connection waveguide connecting the wave splitter / combiner and the optical transversal filter circuit and the monitor port waveguide.

  Here, when actually used as a dispersion compensator, signal light propagates through each of the connection waveguides, so there is a problem of increased loss. However, the monitor port waveguide is used only for grasping the characteristics of an optical branching coupler or the like. Since the signal light does not propagate, there is no problem even if the loss increases as long as the characteristics are not hindered.

  Therefore, in this embodiment, the monitor waveguide is not connected at the intersection of the connection waveguide connecting each optical branching coupler or the connection waveguide connecting each optical multiplexing coupler with the monitor port waveguide. A continuous portion was provided so that the connecting waveguide and the monitoring waveguide were separated from each other.

  FIG. 18 shows a schematic diagram of the intersection. As shown in FIG. 18, at the intersection between the connection waveguide 41 and the monitor waveguide 42, a discontinuous portion 43 is provided in the monitor waveguide 42, and an end surface between the discontinuous portions 43 of the monitor waveguide 42. Was made parallel to the connection waveguide 41.

  Thereby, since there is no contact portion with the monitoring waveguide in each path of the signal light that is input from the input / output port 109 and is output again from the input / output port 109, an increase in loss can be suppressed.

  However, depending on the size of the discontinuous part, some loss may occur in the connecting waveguide, or the loss of the monitoring waveguide will be too large, so the relationship between the size of the discontinuous part and the loss at the intersection of each waveguide I investigated.

  FIG. 19 shows the relationship between the vertical distance from the center line of the connection waveguide to the end face of the discontinuous portion (G in FIG. 18; hereinafter referred to as gap G) and the loss of the connection waveguide and the monitoring waveguide. The case of 45 degrees and 63 degrees was examined. For comparison, a case without a discontinuous portion was also examined and plotted in FIG. 19 with a gap G = 0 μm. FIG. 19 shows the measurement with a waveguide having 24 points of intersection as shown in FIG.

  From FIG. 19, it can be seen that for the connection waveguide, the loss is significantly reduced when the gap G = 4.65 μm, and the loss becomes almost zero at 6.4 μm or more. On the other hand, for the monitoring waveguide, the loss increases as the gap G increases, and a loss of about 20 dB occurs when the gap G = 8.1 μm. However, when considering the loss per intersection, the loss is 1 dB / point or less, and the maximum number of intersections in the path used for characteristic measurement is 6 in the case of FIG. 17 (for example, port 112 to port 212). Is 6 dB or less, so it is a loss without any problem for the purpose of grasping characteristics. Therefore, it can be considered that it can be used without problems if G is in the range of about 4.65 to 8.1 μm.

  From this result, a discontinuous portion with a gap G = 6.4 μm was provided in the monitoring waveguide at the intersection of the optical transversal filter circuit shown in FIG.

  A tunable dispersion compensator using the optical transversal filter circuit designed as described above was produced and evaluated in the same manner as in the first embodiment. In the evaluation, the measurement using the monitor port is performed by inputting only the TE polarized light using the polarizer and the polarization maintaining fiber as in the first embodiment, and the dispersion characteristic is measured through the circulator. A single mode fiber was connected to the cable.

  In addition, although the loss at the time of measurement using a monitor port increased by 6 dB at maximum compared with the case of Example 1, it did not become a hindrance in grasping the relationship between an energization amount and a branching ratio or a phase shift.

  FIG. 20 shows the group delay spectrum thus measured, FIG. 21 shows the loss spectrum, and FIG. 22 shows a graph showing changes in the pass bandwidth and chromatic dispersion with respect to the coefficient ε. From these graphs, when the coefficient ε is ± 2.0, the chromatic dispersion is about ± 105 ps / nm, the passband is about 0.6 nm, the in-band minimum loss is about 7.7 dB, and the in-band loss fluctuation is within 1.2 dB. was gotten.

  For comparison, when a tunable dispersion compensator having no discontinuous portion at the intersection was fabricated, it was found that the in-band minimum loss was about 8.8 dB, which was about 1.1 dB higher. This is due to loss at the intersection.

  In addition, this invention is not limited to the said embodiment example, Various aspects can be taken.

  For example, the number of taps of the optical transversal filter that forms the dispersion compensator is not particularly limited, and is appropriately set.

  In addition, the shape of the intersection is not limited to that shown in FIG. 17, for example, as shown in FIG. 23a, the boundary angle between the monitoring waveguide 42 and the discontinuous portion 43 is not parallel to the connection waveguide 41, As shown in FIG. 23b, the shape in which a part in the width direction and the depth direction of the monitoring waveguide 42 and the connection waveguide 41 are in contact with each other, or the discontinuous portion only on the side surface of the connection waveguide 41 as shown in FIG. The shape can be appropriately selected according to the loss required for the connection waveguide 41 and the monitoring waveguide 42.

  In the above embodiment, the quartz optical waveguide manufactured by the FHD method and the RIE method is used. However, the present invention is not limited to this. Various waveguide materials such as polymers and semiconductors, chemical vapor deposition, and the like are used. Various waveguide pattern forming methods such as a method, various film forming methods such as a sputtering method, a vapor deposition method, and a coating method, a lift-off method, wet etching, stamping molding, extrusion molding, and injection molding can be appropriately selected.

  In the above embodiment, the thin film heater using the thermo-optic effect is used as the branching ratio adjusting means and the phase adjusting means. However, the invention is not limited to this, and the electro-optic effect, the acousto-optic effect, and the magneto-optic effect are not limited thereto. Various adjustment means using the above can be selected.

  In the above embodiment, the target characteristic having a quadratic function-like phase characteristic is used. However, the present invention is not limited to this. Various function shapes such as a sine function shape, and the transmission path to be compensated for dispersion are used. For example, the inverse characteristic of the dispersion characteristic can be used as it is as the target characteristic.

  In the above embodiment, the FSR is set to 100 GHz. However, the present invention is not limited to this, and can be set as appropriate in view of the required dispersion compensation amount and band.

The schematic diagram which shows the circuit structure of the variable dispersion compensator by a 1st Example. The schematic diagram which shows the structure of an optical branch coupler and an optical waveguide coupler. The characteristic view which shows the target characteristic used in the 1st Example. The characteristic view which shows the group delay spectrum calculation result of the variable dispersion compensator in a 1st Example. The characteristic view which shows the loss spectrum calculation result of the variable dispersion compensator in a 1st Example. The characteristic view which shows the calculation result of the chromatic dispersion change with respect to the coefficient (epsilon) in a 1st Example. The characteristic view which shows the loss change measurement result with respect to the energization amount of the optical branching coupler VC2-1 in a 1st Example. The characteristic view which shows the result of having converted the loss change with respect to the energizing amount of the optical branching coupler VC2-1 in a 1st Example into the branching ratio change. The characteristic view which shows the interference spectrum change measurement result between phase shifters PS-4 and PS-5 with respect to the energization amount of phase shifter PS-4 in a 1st Example. The characteristic view which shows the measurement result of the interference spectrum peak wavelength change between phase shifter PS-4 and PS-5 with respect to the energization amount of phase shifter PS-4 in a 1st Example. The characteristic view which shows the interference spectrum between each adjacent phase shifter at the time of the no electricity supply in a 1st Example. The characteristic view which shows the interference spectrum between each adjacent phase shifter after the error correction in a 1st Example. The characteristic view which shows the group delay spectrum measurement result of the variable dispersion compensator in a 1st Example. The characteristic view which shows the loss spectrum measurement result of the variable dispersion compensator in a 1st Example. The characteristic view which shows the chromatic dispersion change characteristic with respect to the coefficient (epsilon) in a 1st Example. The schematic diagram which shows the system application example using the polarization diversity of the variable dispersion compensator by a 1st Example. The schematic diagram which shows the circuit structure of the variable dispersion compensator by a 2nd Example. The schematic diagram which shows the cross | intersection part of the variable dispersion compensator by a 2nd Example. The characteristic view which shows the cross loss change characteristic with respect to the gap G from the connection waveguide center to the discontinuous part end surface of the waveguide for monitoring. The characteristic view which shows the group delay spectrum measurement result of the variable dispersion compensator in a 2nd Example. The characteristic view which shows the loss spectrum measurement result of the variable dispersion compensator in a 2nd Example. The characteristic view which shows the chromatic dispersion change with respect to the coefficient (epsilon) in a 2nd Example. The schematic diagram which shows the other modification of the discontinuous part shape of the waveguide for a monitor. The schematic diagram which shows an example of the variable dispersion compensator using the conventional lattice filter. The block diagram which shows an example of the conventional transversal filter.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Optical waveguide 12 Optical waveguide 11 '* 12' Connection waveguide 13 * 14 Directional coupler 15 Thin film heater 16 Heat insulation groove | channel 21 Circulator 22 Polarization splitter / combiner 23 Polarization holding fiber 24 Variable dispersion compensator 31 Waveguide type polarization Wave splitter / combiner 32 Half-wave plate 41 Connecting waveguide 42 Monitoring waveguide 43 Discontinuous portion 104 Optical input port 109 Input / output port 204 Optical output port g _n Tap coefficient G Gap PS Phase shifter VCx-x Optical branching coupler VCy- y optical multiplexing coupler d optical delay line k optical multiplexer ε coefficient ψ1-ψn phase shifter

Claims (7)

An optical waveguide circuit formed on the substrate;
The optical waveguide circuit is:
A multi-stage optical branching coupler having a tree structure in which an optical branching coupler is arranged at each node. The tree structure has one optical branching coupler corresponding to the root node as the first stage and is input to the one optical branching coupler. A multi-stage optical branch coupler having a structure in which light is branched and output, each of the branched and output lights is input to a separate optical branch coupler of the next stage, and the number of branches increases as the stage progresses;
A plurality of optical delay lines that respectively input the light output from the final stage of the multistage spectral coupler, and output the propagation light by delaying the propagation time by a set time,
An optical multiplexing coupler is a multi-stage optical multiplexing coupler having an inverted tree structure arranged at a position symmetrical to the optical branching coupler with the optical delay line in between, and the inverted tree structure has an optical input end included in each of the optical input couplers. A plurality of optical multiplexing couplers connected to the optical delay line in a one-to-one relationship are used as the first stage, and the light input to the first stage optical multiplexing coupler is combined and output, and each of the combined output lights is output. Is input to a separate optical multiplexing coupler in the next stage, the number of branches decreases as the stage progresses, and a multistage optical multiplexing coupler that outputs light from one optical multiplexing coupler in the final stage;
The optical delay lines are arranged in parallel with each other at an interval, and are formed so that the length increases in order from one end side to the other end side in the juxtaposed direction, thereby constituting an optical transversal filter circuit. One or more optical branching couplers are provided with optical branching ratio adjusting means capable of changing the optical branching ratio, and one or more optical multiplexing couplers are provided with optical multiplexing ratio adjusting means capable of changing the optical multiplexing ratio, Each of the optical delay lines of the number or more is provided with an optical phase adjusting means capable of changing the phase of propagating light, and each of the optical branching coupler and the optical multiplexing coupler arranged at symmetrical positions with the optical delay line interposed therebetween . The intensity ratio of the branched light and the intensity ratio of the multiplexed light are adjusted to be equal to each other , and input from the optical input terminal (optical input port) of the first-stage optical branching coupler and the final-stage optical multiplexing coupler Optical output end (optical output port) Pass wavelength characteristic of the light output is adjusted to have a desired group delay characteristic in a predetermined wavelength passband,
On the substrate, the light input from the common port is separated into two orthogonal polarization mode lights and output from the first and second polarization ports, and the first and second polarization ports Is integrated with a waveguide-type polarization splitter / combiner that synthesizes two orthogonally polarized light beams input from each other and outputs them from the common port, and the first or second polarization port has a half-wave plate Is inserted, and the first polarization port is connected to the optical input terminal of the first-stage optical branching coupler, and the second polarization port is connected to the optical output terminal of the final-stage optical multiplexing coupler. A variable dispersion compensator characterized by being connected.   2. The multistage optical branching coupler or multistage optical multiplexing coupler is connected to at least one monitoring waveguide for monitoring the branching ratio of each stage of optical branching coupler or optical multiplexing coupler. The variable dispersion compensator as described.   Each stage of the optical branching coupler is composed of a two-input type optical coupler. Of the two optical input terminals of the optical coupler, the optical input terminal of the tunable dispersion compensator or the optical output terminal of another optical branching coupler. The variable dispersion compensator according to claim 2, wherein an optical input terminal not used for connection is connected to the monitoring waveguide.   The optical multiplexing coupler at each stage is composed of a two-output type optical coupler. Of the two optical output terminals of the optical coupler, the optical output terminal of the tunable dispersion compensator or the optical input terminal of another optical multiplexing coupler. 4. The tunable dispersion compensator according to claim 2, wherein an optical output end not to be connected is connected to the monitoring waveguide.   One or more monitoring waveguides intersect the connecting waveguide connecting each optical branching coupler or each optical multiplexing coupler or the connecting waveguide connecting the polarization splitter / combiner and the optical transversal filter circuit. 5. The variable dispersion compensator according to claim 2, wherein a discontinuous portion is provided in the monitoring waveguide at the intersection.   6. The tunable dispersion compensator according to claim 5, wherein an end face of the monitoring waveguide in contact with the discontinuous portion is parallel to the connecting waveguide.   When the optical delay line array number is 0, 1 and 2 in order from the shortest optical delay line number, and the longest optical delay line number is N-1 (N is an integer of 3 or more). The relationship between the array number of the optical delay line and the light intensity and phase passing through each optical delay line is centered on the optical delay line number of N / 2 when N is an even number, and (N− 7. The variable dispersion compensator according to claim 1, wherein the tunable dispersion compensator is formed symmetrically about an optical delay line array number of 1) / 2.

Images