JP2011069924A - Quadrature phase shift keying (qpsk) modulator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a QPSK modulator that maintains a phase difference between the I phase and the Q phase to π/2. <P>SOLUTION: The QPSK modulator includes: a means for splitting incident light into two beams; a first phase modulating means for modulating the phase of one of the branched beams; a second phase modulating means for modulating the phase of the other branched beam; a phase shifting means disposed in a preceding stage or a succeeding stage of the first or second phase modulating means; a means for coupling the beams that have respectively passed through the first phase modulating means and the second modulating means, and outputting the coupled beams as an optical signal; a means for converting a part of the outputted optical signal into an electric signal; a means for detecting a peak value of the amplitude of the optical signal from the electric signal and outputting a peak detection signal; a means for generating a dither signal; a phase comparing means for comparing phases of the peak detection signal and of the dither signal and generating a deviation signal in accordance with the deviation from the phase shift amount π/2 in the phase shifting means; and a feedback control means for superposing the dither signal on the deviation signal and feedback-controlling the phase shifting means to adjust the phase shift amount to π/2. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an optical QPSK modulator. In particular, the present invention relates to quadrature control of a QPSK signal required for modulation of an optical QPSK signal in ultrahigh-speed optical communication such as 40 Gb / s or 100 Gb / s.

  In the information and communication network, network services such as optical network / NGN (Next Generation Network) and 3G (3rd Generation) / super 3G are being added and enhanced for full-scale development of broadband and ubiquitous services. As a result, services in various fields such as communications and broadcasting, fixed communications and mobile communications will be merged. When such an era of broadband and ubiquitous services arrives, it is indispensable to construct a full IP network infrastructure that will serve as the foundation for service convergence. Until now, broadbandization of optical communication has progressed dramatically by wavelength multiplexing. Recently, the bit rate has been remarkably improved by time division multiplexing in which information is multiplexed on the time axis. Conventional optical communication technology uses binary amplitude shift keying (OOK: On) using the NRZ (Return-To-Zero) or RZ (Return-To-Zero) format with bit rates of 2.5 Gb / s and 10 Gb / s. -Off Keying), and the method of increasing the capacity by wavelength multiplexing was the mainstream. In recent years, modulation / demodulation techniques such as duobinary transmission, CSRZ (Carrier-Suppressed Return-To-Zero), optical DPSK (Differential Phase Shift Keying), and optical DQPSK (Differential Quadrature Phase Shift Keying) have begun to be applied to optical communication. The bit rate has been increased from 10 Gb / s to 40 Gb / s. Among these, the DQPSK modulation / demodulation format that is particularly advantageous for improving the bit rate is recognized as a promising candidate for future 40 Gb / s optical communication. The DQPSK code converts digital electric signals (0, 0), (0, 1), (1, 0), (1, 1) into four phase changes (0, π / 2, π, 3π / 2). It is a quaternary code to be assigned. Therefore, when the same symbol rate is used, DQPSK can allocate twice as much information as NRZ and DPSK, and the spectral efficiency is doubled. And polarization mode dispersion.

  Furthermore, recently, OIF (Optical Internetworking Forum), which is a global standard organization in the communication industry, has been active in the next generation of ultra-high-speed communication technology. OIF is actively discussing the standardization of communication systems / device technologies that will realize the next generation of 100Gb / s optical communications. As a specific communication method, a DP-QPSK (Dual Polarization-Quadrature Phase Shift Keying) method is considered promising toward standardization of OIF. The DP-QPSK system basically uses a four-level phase modulation system as in DQPSK. In DQPSK, a data signal is assigned to a phase change from the previous bit, whereas in QPSK, a data signal of 0 or 1 is assigned to the absolute value of the phase itself. Therefore, on the receiving side, in DQPSK, demodulation is performed through a 1-bit delay interferometer to correlate with the previous bit, while in QPSK, a local signal light source serving as an absolute reference of the phase is provided, and a 90 ° hybrid Demodulation is performed by using a coupler to correlate local light and received signal light. DP-QPSK increases the multiplicity and expands the total transmission capacity by polarization multiplexing two-channel QPSK signals into two orthogonally polarized waves. Ultimately, DP-QPSK secures double transmission capacity by quaternary coding of QPSK and further double transmission capacity by polarization multiplexing when the symbol rate is about 25 Gb / s. As a result, a total transmission capacity of 100 Gb / s or more is achieved.

  As described above, the optical QPSK modulation technique which is a quaternary phase modulation system is an important basic element in ultrahigh-speed optical communication such as 40 Gb / s and 100 Gb / s. In general, a QPSK modulator using lithium niobate (LiNbO3) is used to modulate an optical signal with QPSK. In DQPSK modulation, processing corresponding to 1-bit delay is precoded on a 2-channel high-speed main signal and high-speed modulation is performed on the I-phase modulation unit and the Q-phase modulation unit. The I-phase modulation unit and the Q-phase modulation unit are modulated at high speed. Thus, except for the presence or absence of logical processing of the main signal, the hardware part for driving the QPSK modulator is the same for both DQPSK and QPSK. A high-speed main signal of 2 channels is amplified to an amplitude corresponding to 2Vπ of the lithium niobate modulator using a high-speed amplifier, and the I-phase modulation unit and the Q-phase modulation unit are driven. In addition, the phase relationship between the I phase and the Q phase needs to be orthogonal. The QPSK modulator has an externally applicable electrode for controlling the quadrature phase between the I-phase modulator and the Q-phase modulator, and maintains the orthogonality between the I-phase and Q-phase by performing bias control by appropriate feedback. Yes.

  Here, a drift phenomenon in which the driving voltage changes widely over a long time is known for lithium niobate. For this reason, a bias control circuit for suppressing the drift phenomenon is essential. The I-phase / Q-phase modulation unit has the same configuration as that of a normal BPSK modulator. The I-phase / Q-phase modulation unit superimposes the dither signal on the phase modulation unit, monitors a part of the modulator signal light output, and performs synchronous detection with the signal phase of the oscillator that is the source of the dither signal. Can be detected. Therefore, the I-phase / Q-phase modulation unit can always suppress the drift phenomenon by configuring a feedback loop that applies a DC bias according to the deviation of the bias voltage. On the other hand, since there is no correlation between the I-phase and Q-phase signals with respect to the quadrature phase drift phenomenon, the dither signal is simply superimposed on the quadrature phase section to perform synchronous detection on the optical power monitor of the modulator signal light output Even if is applied, an error signal cannot be generated. That is, the detection result by the optical power monitor does not change with respect to the phase of the quadrature phase portion. Therefore, it cannot be determined whether the normal dither method and the optical power monitoring method are orthogonal or deviated from orthogonal.

  Regarding quadrature phase control of a QPSK modulator, it is known that quadrature control electrodes are provided at two locations on the I-phase side and Q-phase side, and each electrode superimposes a dither signal orthogonal to each other and performs synchronous detection using an optical power monitor. (Patent Document 1). As shown in FIG. 1, a conventional optical QPSK modulator is provided with an electrode 1 (6) and an electrode 2 (7) as electrodes relating to a quadrature phase. The optical QPSK modulator splits the light from the semiconductor laser 1 into two and inputs it to the data modulator 5. One of the branched lights passes through the phase shift unit 2. In FIG. 1, the phase shift unit 2 is arranged in the preceding stage of the data modulation unit 5, but an embodiment in which it is arranged in the subsequent stage as shown in FIGS. Further, the optical QPSK modulator branches the signal from the dither signal oscillator 14 into two, and applies one to the electrode 1 (6) and the other to the electrode 2 (7) via the phase shifter 8. The optical signal output from the modulator is received by a low-speed PD (photo detector) 10, phase comparison is performed (12), and the result is feedback-controlled to the phase shift unit 2. By setting the phase difference between the dither signals to an odd multiple of 90 degrees by the phase shifter 8, the correlation between the I phase and the Q phase can be estimated via the dither signal. As a result, a normally used synchronous detection technique can be applied, and high-sensitivity orthogonal control can be realized. Each light that has passed through the data modulator 5 is combined in the intensity modulator 9 and output as an optical signal.

JP 2007-043638 A JP 2007-082094 A

  However, in the conventional control method in which two lines of dither signals orthogonal to the quadrature phase shift control unit are applied, it is necessary to provide two electrodes on the QPSK modulator that can control the quadrature phase. This is not desirable for making the QPSK modulator compact. Further, many main QPSK modulators have one electrode for quadrature phase control, and in this case, it is essentially impossible to perform quadrature control with this control method. Furthermore, this control method requires one additional phase shifter in the feedback loop. Moreover, this phase shifter must be adjusted so that the phase between the two dither signals is exactly 90 degrees. The need for a new additional phase shifter and the need for precise adjustment are essential problems in terms of price, size, and power consumption in realizing an optical QPSK modulator.

  The present invention has been made in view of such a problem, and an object of the present invention is to provide a special electrode configuration, a phase shifter that causes an increase in cost and an increase in size in quadrature phase control of an optical QPSK modulator. The purpose is to solve the problem that parts are required. Another object of the present invention is to realize a QPSK modulator capable of keeping the phase difference between the I phase and the Q phase at π / 2.

  In order to achieve such an object, a QPSK modulator of the present invention includes a branching unit that splits input light into two, a first phase modulation unit that phase-modulates one of the branched light, and the other branched Second phase modulation means for phase-modulating the light, phase shift means provided in the preceding stage or subsequent stage of any of the first phase modulation means and the second phase modulation means, and the first phase Coupling means for coupling the light that has passed through the modulation means and the second phase modulation means and outputting as an optical signal; and photodetecting means for converting a part of the optical signal output from the coupling means into an electrical signal; Detecting a peak value of the amplitude of the optical signal from the electrical signal and outputting a peak detection signal; a dither signal generating means for generating a dither signal; and a phase of the peak detection signal and the dither signal. Compare The phase comparison means for generating a deviation signal corresponding to the deviation from the phase shift amount π / 2 in the phase shift means, and the dither signal is superimposed on the deviation signal so that the phase shift amount is π / 2. Feedback control means for performing feedback control on the phase shift means so that

  According to the present invention, it is possible to control the quadrature phase using a general-purpose QPSK modulator without requiring a QPSK modulator having a special electrode structure. Further, an additional phase shifter and adjustment of the phase shifter can be omitted, and an inexpensive, small, and low power consumption QPSK transmitter can be realized.

  The quadrature phase control method of the present invention can apply a commonly used synchronous detection technique, and can form a stable and highly sensitive feedback loop without using a special control method.

  It is obvious that the phase control method of the present invention is suitable for controlling a single QPSK modulator, but further, a signal format (for example, NRZ-DQPSK, RZ-DQPSK, Is a multi-level code such as 8PSK and 16PSK using QPSK codes as an applied form, and a multi-level QAM code, OFDM code, and DP-QPSK, etc. It can be applied to quadrature control.

It is a block diagram of the conventional optical QPSK modulator bias control. It is a basic block diagram of optical QPSK modulator quadrature phase bias control shown in Embodiment 1 of the present invention. It is a figure which shows the principle of optical QPSK modulator quadrature phase bias control shown in Example 1 of this invention. It is a basic block diagram of the optical QPSK modulator quadrature phase bias control shown in Embodiment 2 of the present invention. It is a block diagram of optical QPSK modulator bias control shown in Example 3 of the present invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 2 shows a basic configuration of the QPSK modulator 17 according to the first embodiment of the present invention. FIG. 2 also includes an optical branch 18, a photodetector 19, a peak hold circuit 20, a phase comparison 21, a dither signal oscillator 22, and a drive circuit 23.

  First, the optical signal from the output part of the QPSK modulator 17 composed of lithium niobate is branched (18), and the optical detector 19 converts the optical signal into an electrical signal. The peak hold circuit 20 generates an output that maintains the maximum value (peak value) of the amplitude with respect to the monitor electrical signal converted into the electrical signal. The phase of the signal detected in this way and the signal from the dither signal oscillator 22 are compared in phase (21), and a deviation component is extracted. The deviation component obtained by the phase comparison is processed as a DC signal having an appropriate magnitude and time constant, and is given to the quadrature phase control unit of the QPSK modulator by feedback. At that time, the dither signal from the dither signal oscillator 22 is superimposed on the DC signal. The dither signal superimposed on the DC signal is converted into a driving electric signal suitable for controlling the phase of the lithium niobate by the driving circuit 23 and fed back to the quadrature phase control electrode.

  FIG. 3 shows the principle of quadrature control of the QPSK modulator according to the first embodiment of the present invention. The graph of FIG. 3 shows the result of simulating the state of the peak value of the optical signal based on the constellation. The horizontal axis of the graph represents the phase difference between the IQ modulators, and the vertical axis represents the relative optical signal amplitude. Then, the larger one of the amplitudes (01) and (11) shown in FIG. 3 is the peak value of the amplitude. In the case of an ideal quadrature state with an IQ phase difference of 90 degrees, the constellation (1) forms a square. In this case, the amplitudes of the four values (00), (01), (10), and (11) obtained by synthesizing the IQ signal are degenerated and become equal values, which becomes the peak value of the amplitude. Next, when the IQ phase difference deviates from 90 degrees, the constellation (2) forms a rhombus. Compared to the orthogonal component (amplitude peak value) of the constellation (1) in the orthogonal state, the long axis component is expanded and the short axis component is shortened in the constellation (2). When the IQ phase difference further shifts and approaches 180 degrees (or 0 degrees) as shown in constellation (3), the short axis component disappears and only the long axis component remains. The major axis component at this time is √2 times the diagonal component when orthogonal. From this simulation result, the amplitude peak value of the output signal of the QPSK modulator becomes minimum in the orthogonal state (constellation (1)), and the amplitude peak value increases when the modulator orthogonal bias shifts from the orthogonal state (constellation). (2) (3)). Therefore, it is possible to estimate the state of the quadrature phase by detecting the amplitude peak value of the QPSK modulator output signal.

  Furthermore, it should be noted that when the graph of FIG. 3 is viewed from the viewpoint of control, it has a characteristic that the amplitude peak value is minimized in the orthogonal state. In other words, this is achieved by superimposing a low-speed dither signal on the quadrature control electrode, monitoring the modulator output, detecting the amplitude peak value of the dither signal component, and comparing the phase with the original dither signal waveform. This means that a deviation signal corresponding to the deviation from the state can be extracted. When the quadrature phase control bias matches the quadrature state, the deviation signal is zero. This operation is the same as the synchronous detection by dithering generally used.

  What is more characteristic and important is that the orthogonal state becomes an extreme value having a sharp inclination. That is, even if the phase is slightly shifted from the orthogonal state, the amplitude peak value reacts sensitively and shows a large value. Having such a sharp bottom characteristic means that the sensitivity of deviation detection is high, which is a great merit in converging control more stably. This means that it is not necessary to extract intensity frequency components (symbol rate frequency components) by combination with other high sensitivity means, for example, an RZ intensity modulator. For this reason, high-sensitivity quadrature phase control can be realized by the present invention alone.

  2 is compared with the operation of the circuit configuration of FIG. 2, it can be seen that the amplitude peak value of the signal component can be obtained by passing the monitor signal of the modulator output through the peak hold circuit 20. From the graph of FIG. 3, since the amplitude peak value is minimum in the orthogonal state, a deviation signal can be obtained by comparing the phase of the output result of the peak hold circuit 20 with the dither signal waveform. Note that phase comparison is a commonly used technique, which is obtained by multiplying a target signal waveform and a reference signal waveform.

  It is important to note that the photodetector operating band and the IQ phase signal frequency band at least partially overlap. As shown in the graph of FIG. 3, the IQ phase is modulated at a high speed by the symbol frequency, and the amplitude peak value cannot be detected unless at least a part of the signal frequency component is detected. Even if the peak hold is applied, the behavior as shown in the graph is not shown. Specifically, as shown in Patent Document 2, when 43 Gb / s DQPSK modulation is assumed, the symbol frequency is 21.5 GHz, and the operation band of the photodetector is 3 dB or more when the cutoff frequency is 100 MHz or more. It should be good. Operating the photodetector at about 100 MHz can be adequately handled by ordinary circuit components and circuit design. Further, if the operating band of the peak hold circuit is set to about this cut-off frequency, the amplitude peak value can be detected more efficiently. Furthermore, the dither signal frequency should be sufficiently lower than that of the peak hold circuit, and it should be about several KHz to several MHz for an operating band of about 100 MHz. This frequency is generally the same as the dither control frequency of the lithium niobate intensity modulator.

  The optical branching unit for monitoring the output of the modulator has been described as being externally attached to the modulator. However, in some cases, a tap is incorporated in the output unit of the QPSK modulator. In this case, the tap is used. It is preferable to reduce the size. In addition, it is very effective to perform proportional / integral / derivative (PID) calculation processing in order to stabilize the control loop in the phase comparison unit 21, and as long as the feedback converges stably, a part of the calculation processing and combinations thereof It is self-evident in control design that it can be implemented.

  As described above, if the quadrature phase control method of the present invention is used, a control circuit can be realized with only ordinary synchronous detection means, and the phase difference between dither signals performed by the conventional control method (FIG. 1) can be reduced by 90 °. No separate phase adjustment process is required. PID parameter adjustment is required as before to improve the stability of the control loop, but individual differences such as the dynamic characteristics of the modulator and electronic components are small in the frequency range of about 100 MHz. There is no need to make individual adjustments. In addition, it is necessary to optimize the time constant (capacitor capacity) of the peak detection circuit in order to increase the control sensitivity, but the bit rate of the main signal does not change by more than one digit at about 20 Gbps, so it is optimized once Once adjusted, there is no need to make individual adjustments.

  Therefore, the quadrature control method of the present invention eliminates the need for individual adjustment operations thereafter once the closed-loop parameters are optimized. In other words, once the design is finalized, the control method is essentially autonomous.

  FIG. 4 shows a specific circuit configuration of quadrature control of the QPSK modulator 24 according to the second embodiment of the present invention. 4 includes a photodetector 26, a current detection resistor 27, and an amplifier 28. The peak hold circuit includes a diode 31, a resistor 32, and a capacitor 33. After the photocurrent is detected by the photo-detector 26, an AC signal is taken out by the capacitive coupling 29 to remove the DC component, amplified by the buffer amplifier 30, and then input to the peak hold circuit (31, 32, 33). The output of the peak hold circuit is appropriately amplified (34), and only the dither signal component is extracted by the band pass filter 35. The signal extracted by the peak hold circuit and the band pass filter 35 is multiplied by the original dither signal waveform (37), and converted to a DC component by the low pass filter 39 to generate a deviation signal. The deviation signal is subjected to proportional / integral / derivative control generally used in feedback control (40) to generate a DC bias signal. The dither signal waveform is added to the DC bias signal having an amplitude corresponding to the phase shift of the quadrature portion generated in this way (41), and converted into an appropriate voltage signal by the driver 42 that drives the lithium niobate, for quadrature phase control. Apply to electrode. The operating bands of the photodetector, peak hold circuit, and dither signal oscillator are the same as in the first embodiment.

  All of these circuit elements can be constituted by analog circuits. Also, the signal after the monitor unit is appropriately converted from analog to digital, and a peak hold circuit, a band pass filter unit, a multiplier circuit, a low pass filter, a proportional / integral / differential calculation processing unit, a dither signal oscillator, a part or all of the adder circuit A configuration in which a digital circuit such as a CPU or FPGA, digital-analog conversion, and input to a driver to drive a modulator bias electrode is suitable for miniaturization.

  Each component and circuit configuration of the embodiment shown here are derived from the viewpoint of optimum design, but some of them are omitted or a circuit is omitted for convenience of implementation such as downsizing and low power consumption. Modifications such as changing the order of the parts are not particularly limited.

  FIG. 5 shows a specific circuit configuration for performing bias control of the entire QPSK modulator according to the third embodiment of the present invention. The quadrature bias control circuit is similar to the circuit configuration shown in the second embodiment, and includes a resistance detection circuit 46, 47, 48, a capacitive coupling 49, a buffer amplifier 50, a diode 51, a resistor 52, and a capacitor 53. The circuit includes a band pass filter 55, amplifiers 54 and 56, a multiplier 57, a low pass filter 59, a proportional / integral / differential operation circuit 60, a dither signal oscillator 58, and an adder circuit 61. As a result, the quadrature phase is controlled to the optimum value as described in the first embodiment.

  Independent of the quadrature bias control circuit, bias control of the BPSK modulation unit for I modulation and the BPSK modulation unit for Q modulation is configured. The bias control of the I modulator and the Q modulator is a general configuration in which a dither signal component is detected by an optical power monitor and the phase is compared. However, when monitoring the output of the modulator, it is desirable that the monitor of the I modulator and the Q modulator and the monitor of the quadrature phase control are independent. This is because the band of the monitor unit of the I modulator and the Q modulator is as low as the dither frequency, while the band of the monitor unit of the quadrature phase control is specifically described in the first embodiment. This is because a certain degree of broadband characteristics are required to detect a part of the symbol frequency.

  The I modulator bias control, the Q modulator bias control, and the quadrature phase bias control may be configured so that the respective controls are always operated at the same time, or may be configured to repeat each control in a time division manner in a time division manner. In the case of simultaneous control, it is necessary to change the dither frequency used for each bias control. It is also preferable to incorporate a band-pass filter in each control to suppress crosstalk. On the other hand, in the case of time division control, the dither frequency may be the same. Also, the dither signal oscillator, multiplier, low-pass filter, proportional / integral / differential calculation processing circuit, and adder circuit are shared, and the deviation signal input and the analog output to the driver for driving each electrode are switched and digital. The configuration of switching by selecting the output port of the circuit is convenient for further miniaturization.

  In the first to third embodiments, a configuration example using a modulator using lithium niobate as a QPSK modulator is shown, but the material of the modulator is not limited to this, and indium phosphide (InP) is also applicable. Furthermore, the present invention is chip-chip connected by a PLC (Planar Lightwave Circuit), uses lithium niobate or indium phosphine for the high-speed modulation part, and uses a quartz part for controlling the branching part and bias of the interferometer. The present invention can also be applied to a modulator composed of a system optical waveguide. However, in this case, the driver is not driven by voltage like the lithium niobate using the electric field effect, but is driven by electric power to the heater using the thermo-optical effect of the silica-based optical waveguide. Furthermore, since the operating band of the thermo-optic effect of the silica-based optical waveguide is about 100 Hz, the dither frequency is 100 Hz or less.

  In the above embodiments, the configuration of a single QPSK modulator is shown. However, the data modulation unit is precoded to operate as a DQPSK modulator, or the RZ modulator is driven using a data modulation clock component. It is also possible to apply to RZ-DQPSK modulation by arranging the DQPSK modulator before or after the DQPSK modulator.

  The phase control method and configuration of an optical QPSK transmitter according to the present invention can be used for an optical communication device used in an optical communication network or the like.

1 Semiconductor laser (LD: laser diode)
2 Phase shift unit 3 First arm 4 Second arm 5 Data modulation unit 6 First electrode 7 Second electrode 8 Phase shifter 9 Intensity modulator 10, 10 ′ Low-speed light detector (PD: photo detector)
11 Band-pass filter (BPF)
DESCRIPTION OF SYMBOLS 12 Phase comparator 13 Multiplier 14 Low frequency signal generator 15 Clock signal generator 16 Drive signal generator 17, 24, 43 QPSK modulator 18, 25, 44, 45 Optical branch 19 Photo detector 20 Peak hold circuit 21 Phase Comparison 22, 38, 58, 67 Dither signal oscillator 23 Drive circuit 26, 46 Medium speed range (~ 100MHz) photodetector 27, 47, 64 Resistance 28, 30, 48, 50 Medium speed range (~ 100MHz) amplifier 29, 49 Capacitor 31, 51 Diode 32, 52 Resistor 33, 53 Capacitor 34, 36, 54, 56, 65 Amplifier 35, 55 Band pass filter (BPF)
37, 57, 66 Multiplier 39, 59, 68 Low-pass filter (LPF)
40, 60, 69 Proportional Integral Derivative (PID) processing circuit 41, 61, 70, 72 Adder 42, 62, 73 Driver

Claims (3)

A branching means for splitting the input light into two;
First phase modulation means for phase-modulating one of the branched lights;
Second phase modulation means for phase-modulating the other branched light;
A phase shift means provided at a preceding stage or a subsequent stage of any of the first phase modulating means and the second phase modulating means;
Coupling means for coupling the light that has passed through the first phase modulation means and the second phase modulation means and outputting as an optical signal;
Photodetection means for converting a part of the optical signal output from the coupling means into an electrical signal;
Peak detection means for detecting a peak value of the amplitude of the optical signal from the electrical signal and outputting a peak detection signal;
Dither signal generating means for generating a dither signal;
A phase comparison unit that generates a deviation signal according to a deviation from the phase shift amount π / 2 in the phase shift unit by comparing phases of the peak detection signal and the dither signal;
A QPSK modulator comprising: feedback control means for superimposing the dither signal on the deviation signal and feedback-controlling the phase shift means so that the phase shift amount becomes π / 2.   The QPSK modulator according to claim 1, wherein the first phase modulation means and the second phase modulation means are made of lithium niobate.   2. The QPSK modulator according to claim 1, wherein the first phase modulation means and the second phase modulation means are made of indium phosphide.

Images