Optical modulation device and bias voltage control method

Abstract

An optical modulation device including: bias power supplies that output a signal having a bias voltage corresponding to the null point of an optical modulation unit to the optical modulation unit; and synchronous detection circuits that determine whether an intensity of a QAM signal at a drift non-occurrence time where no drift occurs in the bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the bias voltage, adjust the bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, and adjust the bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time.

Description

BACKGROUND

1. Field of the Disclosure

The present disclosure relates to an automatic bias voltage control of an optical modulation device for use in multi Quadrature Amplitude Modulation (QAM). In particular, the present disclosure is suitable for the automatic bias voltage control of an optical modulation device that sends a quaternary or larger value QAM signal.

2. Discussion of the Background Art

As a transmission code for use in an optical transmission system, a QAM signal, capable of sending a large amount of optical signals at a low symbol rate has been paid attention. The simplest QAM is quaternary QAM and called Quadrature Phase Shift Keying (QPSK). The present disclosure is applicable to any multi-QAM modulators including QPSK, however for simplicity, a description will be mainly given of a 16 QAM system in the present disclosure. Here, some symbols have a bar placed on characters in figures, however placed before the characters indicates that the symbols have the bar on the characters in the specification.

FIG. 1 shows the configuration of an optical modulation device in the related art. A continuous optical signal input to an IQ optical modulator M is divided into two signals by a first optical coupler 1 and input to the first and second optical modulation units which are an I-component optical modulation unit 2 and a Q-component optical modulation unit 3 (hereinafter first and second optical modulation units 2 and 3, respectively). The first and second optical modulation units 2 and 3 are generally composed of Mach-Zehnder Interferometer (MZI) type optical modulators and have the function of relatively changing an optical phase and optical intensity corresponding to the logic of the first quaternary data signals Data1 and Data1 and that of second quaternary data signals Data2 and Data2. Here, the relationship between the optical phase and optical intensity with respect to the four values of the data signals will be described below. In addition, the first and second bias voltages applied to the first and second optical modulation units 2 and 3, respectively, will be described below.

After the addition of a phase difference θ₃ by an optical phase shifter 4 having an orthogonal bias electrode 101 to which a third bias voltage to be described below is applied, the outputs of the first and second optical modulation units 2 and 3 are multiplexed by the second optical coupler 5 and output as a 16 optical QAM signal. If θ₃ is ±π/2, an ideal waveform can be obtained. It is equivalent to ¼ of a carrier wavelength, and general wavelength is of the order of micrometer, so then the adjustment must be performed really precisely. In addition, since the optical quality of an optical QAM signal is sensitive to an error in the optical phase shifter 4, it is really important to adjust the amount of phase shift induced by the optical phase shifter 4 to an appropriate value.

In general, this adjustment is made in such a manner that a third bias voltage (also called an orthogonal bias voltage) V_bias3 supplied from a third bias power supply 10 to the optical phase shifter 4 is adjusted. The optical phase shifter 4 is arranged on the rear stage of the second optical modulation unit 3 in FIG. 1, however it may be arranged on the rear stage of the first optical modulation unit 2, arranged on those of both the first and second optical modulation units 2 and 3, or arranged on the front stage. Hereinafter, for simplicity, it is assumed that the optical phase shifter 4 is arranged only on the rear stage of the second optical modulation unit 3.

Next, a description will be given of the relationship between the optical phase and optical intensity in the outputs of the first and second optical modulation units 2 and 3 with respect to the four values of the first and second quaternary data signals. As described above, it is general to use MZI type optical modulators as the first and second optical modulation units 2 and 3. The first and second optical modulation units 2 and 3 are driven by the first and second quaternary data signals, respectively. Each of these data signals is a quaternary NRZ Non Return-to-Zero (NRZ) signal. The first driving amplifier 6 amplifies the first quaternary data signal into the two complementary signals, and generates Data1 and Data1. The second driving amplifier 7 amplifies the second quaternary data signal into the two complementary signals, and generates Data2 and Data2.

The respective amplified quaternary data signals Data1 and Data1 are applied to the two arms of the first optical modulation unit 2 via a first driving electrode 61 to generate a phase shift ±φ₁. The respective amplified quaternary data signals Data2 and Data2 are applied to the two arms of the second optical modulation unit 3 via a second driving electrode 71 to generate a phase shift ±φ₂. The values of the phase delays φ₁ and φ₂ change with the four values of the respective data signals. In addition, DC voltages (data bias voltages) V_bias1 and V′_bias1 are generated by a first bias power supply 8, and optical phase shifts +θ₁ and −θ′₁ are further added via a first bias electrode 81. Moreover, DC voltages (data bias voltages) V_bias2 and V′_bias2 are generated by a second bias power supply 9, and optical phase shifts +θ₂ and −θ′₂ are further added via a second bias electrode 91.

Next, the expression of the various voltages described above is defined as follows. The four signal levels of a differential signal (Data1- Data1) generated by the first driving amplifier 6 are expressed as V₀, V₁, −V₁, and −V₀, and it is assumed that V₀>V₁>−V₁>−V₀ is established. Since the optical characteristics of the first optical modulation unit 2 are generally the same as those of the second optical modulation unit 3, the four signal levels of a differential signal (Data2- Data2) generated by the second driving amplifier 7 are also expressed as V₀, V₁, −V₁, and −V₀.

V_bias1, V′_bias1, V_bias2, and V′_bias2 select the null points of the first and second optical modulation units 2 and 3. In other words, the output light of the first and second optical modulation units 2 and 3 is set to extinguish when the differential voltages of the signals generated by the first and second driving amplifiers 6 and 7 are 0. In addition, the maximum amplitudes of the differential outputs of the first and second driving amplifiers 6 and 7 are set so as not to exceed twice the half wavelength voltages V_π of the first and second optical modulation units 2 and 3. Accordingly, 2V_π≧V₀−(−V₀)=2V₀ is established.

Here, a description will be given, with reference to FIG. 1, of the half wavelength voltage V_π of the first optical modulation unit 2. The first optical modulation unit 2 is a MZI type modulator and incorporates two waveguides. It is assumed that a data bias is adjusted to extinguish the output of the first optical modulation unit 2 when both of the voltages Data1 and Data1 applied to the two waveguides are 0. In general, complementary driving signals are applied to the two waveguides, and 2V_x is called the half wavelength voltage V_π of the first optical modulation unit 2 if the optical output of the first optical modulation unit 2 reaches maximum intensity under Data1=V_x and Data1=−V_x. The optical output of the first optical modulation unit 2 reaches the maximum intensity even under Data1=−V_x and Data1=V_x, however the optical phase of the optical output in this case is made different by π compared with the former example. Since the first optical modulation unit 2 changes an optical phase by making use of the property, each of Data1 and Data1 is set to have an amplitude of 2V_x=V_π at the maximum and (Data1- Data1) is set to have an amplitude of 2V_π at the maximum. The second optical modulation unit 3 is configured in the same manner as the first optical modulation unit 2.

Here, in the IQ optical modulator M shown in FIG. 1, each of the driving signal electrodes is configured to apply positive and negative complementary voltages to the two waveguides, and thus four electrodes in total exist. The IQ optical modulator M of this type is called a dual drive type. On the other hand, a single drive type IQ optical modulator M has only two driving signal electrodes. In such a configuration, the first driving electrode 61 simultaneously applies electric fields to the two waveguides inside the first optical modulation unit 2, and the second driving electrode 71 simultaneously applies electric fields to the two waveguides inside the second optical modulation unit 3. With the anisotropy of the four waveguides, the single drive type IQ optical modulator M can realize the same function as that of the dual drive type. Also in this configuration, the quaternary data signals applied to the first and second driving electrodes 61 and 71 are the four voltages V₀, V₁, −V₁, and −V₀, and the amplitude of each of the driving signals is set so as not to exceed twice the half wavelength voltage V_π.

FIG. 2 shows the characteristics of the optical modulation device in a case in which no bias drift occurs. FIG. 2 shows the relationship between the electric field E₁ of the output light of the first optical modulation unit 2 and V₀, V₁, −V₁, −V₀, and V_bias1, and the relationship between the electric field E₂ of the output light of the second optical modulation unit 3 and V₀, V₁, −V₁, −V₀, and V_bias2 is also expressed just like FIG. 2. When the sum of the potentials V₀, V₁, −V₁, and −V₀ of a driving signal and the data bias V_bias1 is shown in the horizontal axis and the electric field E₁ of the output light is shown in the vertical axis, a sine wave is drawn. When the optimum data bias is applied, V₀, V₁, −V₁, and −V₀ are arranged symmetrically to the null point and the electric fields E₁₁, E₁₂, E₁₃, and E₁₄ of the output light generated by V₀, V₁, −V₁, and −V₀ are also arranged symmetrically to the 0 level.

FIG. 3 shows a constellation in a case in which no bias drift occurs. Since θ₃=π/2 is established when optimum V_bias3 is applied to the optical phase shifter 4 and the optical phase of the output light of the first optical modulation unit 2 and that of the second optical modulation unit 3 are kept to be orthogonal to each other, the constellation of the output light of the IQ optical modulator M is one having a lattice pattern as shown in FIG. 3. Here, it is important that the respective symbols of the constellation are arranged symmetrically to the origin. This symmetry is a property common to QAM in addition to 16QAM. Here, since E₁ and E₂ are orthogonal to each other when θ₃=π/2 is kept, an In-Phase component and a Quadrarure-Phase component are sometimes briefly called an I-component and a Q-component, respectively.

The optical power P_total of a 16QAM signal is proportional to the sum of the square of the electric field of each of the symbols of the constellation. P_total is expressed by Equation 1.P_total∝Σ(|E_1K|²+|E_2L|²)  (Equation 1)

Here, K and L have the sum of 1 to 4.

In a case in which V_bias1 and V_bias2 are kept at optimum values even if a bias drift occurs in V_bias3 and θ₃ has a value different from π/2, the total of the optical power P_total does not change as described in detail in Non-Patent Literature 1. Therefore, it is relatively difficult to detect a drift occurring in V_bias3. However, it becomes possible to detect a drift occurring in V_bias3 with the application of dithering to V_bias1 and V_bias2 using the technique of asymmetric bias dithering described in Non-Patent Literature 1.

Next, consideration will be given to a case in which a bias drift occurs in the data bias V_bias1 or V_bias2. In the case of quaternary QAM, i.e., a QPSK signal, since the optical power of modulation light immediately decreases after a bias drift occurs in the data bias V_bias1 or V_bias2, it is relatively easy to detect the bias drift occurring in the data bias V_bias1 or V_bias2. Thus, it becomes possible to detect and correct bias drifts occurring in all the V_bias1, V_bias2, and V_bias3 of the QPSK modulator with the technique of the asymmetric bias dithering described above.

CITATION LIST

Non-Patent Literature

Non-Patent Literature 1: H. Kawakami, E. Yoshida and Y. Miyamoto, “Asymmetric dithering technique for bias condition monitoring in optical QPSK modulator,” Electronics Letters (2010), vol.46, no.6, pp.430-431

However, in the case of a quaternary or higher order QAM signal, the number of symbols in the constellation increases, which causes other problems not seen in QPSK.

FIG. 4 shows the characteristics of the optical modulation device in a case in which a bias drift occurs. Since the optical characteristics of the optical modulation device are affected by the bias drift, the curve of the sine wave shown in FIG. 2 shifts rightward or leftward as a whole. FIG. 4 shows a case in which the curve of the sine wave shifts rightward. The four values of E₁ shift from the positions of large white circles to those of large black circles. The absolute value of E₁₃ increases, on the contrary that of E₁₂ decreases. The absolute value of E₁₄ increases on the contrary that of E₁₁ decreases, though these shifts are smaller than those of E₁₂ and E₁₃.

FIG. 5 shows a constellation in a case in which a bias drift occurs. For simplicity, it is assumed that a drift occurs only in V_bias1 and no drift occurs in other biases (V_bias2 and V_bias3). Some symbols (which are satisfying E₁=E₁₃) are distant from the origin, while other symbols (which are satisfying E₁=E₁₂) are close to the origin. For this reason, it is not certain whether the optical power P_total increases or decreases.

FIG. 6 shows an example of the relationship between the drift of the bias voltage and the intensity of the optical signal. The horizontal axis expresses drift amounts of V_bias1, normalized by V_π of the first optical modulation unit 2. It means that, −0.5 in the horizontal axis expresses that V_bias1 decreases by 0.5 V_π from an optimum voltage. Here, it is assumed that V₀−V₁=V₁−(−V₁) is established and the four signal levels of the differential signal (Data1- Data1) generated by the first driving amplifier 6 are arranged at even intervals. Of the two curves shown in FIG. 6, one shows V₀−V₁=V₁−(−V₁)=0.6 V_π and the other shows V₀−V₁=V₁−(−V₁)=0.4 V_π. P_total becomes maximum when V_bias1 is optimum in the former case, however P_total becomes minimum when V_bias1 is optimum in the latter case.

FIG. 7 shows another example of the relationship between the drift of the bias voltage and the intensity of the optical signal. Here, it is assumed that pre-emphasis is applied to V₀ and V₁ and Equation 2 is established.sin(V₀/V_π×π/2)−sin(V₁/V_π×π/2)=sin(V₁/V_π×π/2)−sin(−V₁/V_π×π/2)  (Equation 2)

In the case of FIG. 6, V₀, V₁, −V₁, and −V₀ are arranged at even intervals, however E₁₁, E₁₂, E₁₃, and E₁₄ are not arranged at even intervals, because of the nonlinear characteristics of the first optical modulation unit 2. In the case of FIG. 7, a more ideal constellation can be obtained since E₁₁, E₁₂, E₁₃, and E₁₄ are arranged at even intervals by utilizing the pre-emphasis.

Of the two curves shown in FIG. 7, one establishes Equation 3,sin(V₀/V_π×π/2)−sin(V₁/V_π×π/2)=sin(V₁/V_π×π/2)−sin(−V₁/V_π×π/2)=0.63  (Equation 3)

and the other establishes Equation 4.sin(V₀/V_π×π/2)−sin(V₁/V_π×π/2)=sin(V₁/V_π×π/2)−sin(−V₁/V_π×π/2)=0.60  (Equation 4)

P_total becomes maximum when V_bias1 is optimum in the former case, however P_total becomes minimum when V_bias1 is optimum in the latter case.

In a case in which V_bias1 and V_bias2 are controlled while monitoring the optical power of a multi-QAM signal with a low speed optical power monitor, it is required to appropriately select whether to maximize or minimize the optical power in consideration of the amplitude of a driving signal and the half wavelength voltage of an optical modulation unit.

Accordingly, in order to solve the above problems, it is an object of the present disclosure to provide a technique for appropriately selecting whether to maximize or minimize an optical power in automatically controlling the bias voltage of an optical modulator.

SUMMARY

To this end, the present disclosure appropriately selects whether to maximize or minimize an optical power in consideration of the amplitude of a driving signal and the half wavelength voltage of an optical modulation unit in automatically controlling the bias voltage of an optical modulator.

The present disclosure provides an optical modulation device, including:

an I-component optical modulation unit that modulates a continuous optical signal for an I-component of a QAM signal;

a Q-component optical modulation unit that modulates a continuous optical signal for a Q-component of the QAM signal;

a phase shift unit that shifts a phase of an optical signal on an input side and/or an output side for the I-component optical modulation unit and/or the Q-component optical modulation unit to adjust a phase difference between an output of the I-component optical modulation unit and an output of the Q-component optical modulation unit after the phase shift to π/2;

a QAM signal generation unit that multiplexes the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to generate the QAM signal;

an I-component data signal output unit that outputs an I-component data signal to the I-component optical modulation unit to cause the I-component optical modulation unit to modulate a continuous optical signal using the I-component data signal;

a Q-component data signal output unit that outputs a Q-component data signal to the Q-component optical modulation unit to cause the Q-component optical modulation unit to modulate a continuous optical signal using the Q-component data signal;

an I-component bias voltage signal output unit that outputs a signal having an I-component bias voltage corresponding to a null point of the I-component optical modulation unit to the I-component optical modulation unit;

a Q-component bias voltage signal output unit that outputs a signal having a Q-component bias voltage corresponding to a null point of the Q-component optical modulation unit to the Q-component optical modulation unit;

an I-component bias voltage signal adjustment unit that

• • determines whether an intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the I-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the I-component bias voltage,
  • adjusts the I-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, and
  • adjusts the I-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time; and

a Q-component bias voltage signal adjustment unit that

• • determines whether the intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the Q-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the Q-component bias voltage,
    • adjusts the Q-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, and
    • adjusts the Q-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time.

In addition, the present disclosure provides an optical modulation device that generates a QAM signal whose order is set to be constant. The optical modulation device includes:

an I-component optical modulation unit that modulates a continuous optical signal for an I-component of the QAM signal;

a Q-component optical modulation unit that modulates a continuous optical signal for a Q-component of the QAM signal;

a phase shift unit that shifts a phase of an optical signal on an input side and/or an output side for the I-component optical modulation unit and/or the Q-component signal optical modulation unit to adjust a phase difference between an output of the I-component optical modulation unit and an output of the Q-component optical modulation unit after the phase shift to π/2;

a QAM signal generation unit that multiplexes the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to generate the QAM signal;

an I-component data signal output unit that outputs an I-component data signal whose amplitude is controlled to be constant to the I-component optical modulation unit to cause the I-component optical modulation unit to modulate a continuous optical signal using the I-component data signal;

a Q-component data signal output unit that outputs a Q-component data signal whose amplitude is controlled to be constant to the Q-component optical modulation unit to cause the Q-component optical modulation unit to modulate a continuous optical signal using the Q-component data signal;

an I-component bias voltage signal output unit that outputs a signal having an I-component bias voltage corresponding to a null point of the I-component optical modulation unit to the I-component optical modulation unit;

a Q-component bias voltage signal output unit that outputs a signal having a Q-component bias voltage corresponding to a null point of the Q-component optical modulation unit to the Q-component optical modulation unit;

an I-component bias voltage signal adjustment unit that maximizes or minimizes an intensity of the QAM signal to adjust the I-component bias voltage to an optimum value; and

a Q-component bias voltage signal adjustment unit that maximizes or minimizes the intensity of the QAM signal to adjust the Q-component bias voltage to an optimum value.

In addition, the present disclosure provides a bias voltage control method, including:

I and Q component optical modulation steps of modulating continuous optical signals for I and Q components of a QAM signal;

a phase shift step of shifting a phase of an optical signal on an input side and/or an output side for the I-component optical modulation step and/or the Q-component optical modulation step and adjusting a phase difference between an output of the I-component optical modulation step and an output of the Q-component optical modulation step after the phase shift to π/2; and

a QAM signal generation step of multiplexing the output of the I-component optical modulation step and the output of the Q-component optical modulation step after the phase shift to generate the QAM signal, wherein the I-component optical modulation step sequentially includes

• • an I-component signal input step of
    • inputting an I-component data signal used to modulate a continuous modulation signal and
    • inputting a signal having an I-component bias voltage corresponding to a null point of the I-component optical modulation step and
  • an I-component bias voltage signal adjustment step of
    • determining whether an intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the I-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the I-component bias voltage,
    • adjusting the I-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, and
    • adjusting the I-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time, and
  • the Q-component optical modulation step sequentially includes
  • a Q-component signal input step of
    • inputting a Q-component data signal used to modulate a continuous modulation signal and
    • inputting a signal having a Q-component bias voltage corresponding to a null point of the Q-component optical modulation step and
  • a Q-component bias voltage signal adjustment step of
  • determining whether the intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the Q-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the Q-component bias voltage,
  • adjusting the Q-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, and
  • adjusting the Q-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time.

According to this configuration, it is possible to appropriately select whether to maximize or minimize an optical power in automatically controlling the bias voltage of an optical modulator.

In addition, the present disclosure provides the optical modulation device, further including:

an I-component calculation unit that calculates a value shown in Equation 5 when a voltage of the I-component data signal is V_mi (where m_i=0, 1, . . . , n_i/2−1 when a number of the I-component data signal level is n_i) and a half wavelength voltage of the I-component optical modulation unit is V_πi;

$\begin{array}{cc}\sum _{m_i=0}^{n_i/2-1}\cos \left(\frac{V_{\mathrm{mi}}}{V_{\pi i}}\pi \right);& \left(\mathrm{Equation}5\right)\end{array}$ and

a Q-component calculation unit that calculates a value shown in Equation 6 when a voltage of the Q-component data signal is V_mq (where m_q=0, 1, . . . , n_q/2−1 when a number of the Q-component data signal level is n_q) and a half wavelength voltage of the Q-component optical modulation unit is V_πq,

$\begin{array}{cc}\sum _{m_q=0}^{n_q/2-1}\cos \left(\frac{V_{\mathrm{mq}}}{V_{\pi q}}\pi \right)& \left(\mathrm{Equation}6\right)\end{array}$

wherein

the I-component bias voltage signal adjustment unit

• • determines that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time when the value shown in Equation 5 and calculated by the I-component calculation unit is negative and
  • determines that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time when the value shown in Equation 5 and calculated by the I-component calculation unit is positive, and

the Q-component bias voltage signal adjustment unit

• • determines that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time when the value shown in Equation 6 and calculated by the Q-component calculation unit is negative and
  • determines that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time when the value shown in Equation 6 and calculated by the Q-component calculation unit is positive.

According to this configuration, it is possible to appropriately select whether to maximize or minimize an optical power in consideration of the amplitude of a driving signal and the half wavelength voltage of an optical modulation unit in automatically controlling the bias voltage of an optical modulator.

In addition, the present disclosure provides the optical modulation device, wherein

V_mi is a constant at all times when a voltage-of the I-component data signal is V_mi (where m_i=0, 1, . . . , n_i/2−1 when a number of the I-component data signal level is n_i),

V_mq is a constant at all times when a voltage of the Q-component data signal is V_mq (where m_q=0, 1, . . . , n_q/2−1 when a number of the Q-component data signal level is n_q),

the I-component bias voltage signal adjustment unit

• • adjusts the I-component bias voltage to maximize the intensity of the QAM signal at all times when a constant set by Equation 7 is negative and
  • adjusts the I-component bias voltage to minimize the intensity of the QAM signal at all times when the constant set by Equation 7 is positive, and

$\begin{array}{cc}\sum _{m_i=0}^{n_i/2-1}\cos \left(\frac{V_{\mathrm{mi}}}{V_{\pi i}}\pi \right)& \left(\mathrm{Equation}7\right)\end{array}$

the Q-component bias voltage signal adjustment unit

• • adjusts the Q-component bias voltage to maximize the intensity of the QAM signal at all times if a constant set by Equation 8 is negative and
  • adjusts the Q-component bias voltage to minimize the intensity of the QAM signal at all times if the constant set by Equation 8 is positive.

$\begin{array}{cc}\sum _{m_q=0}^{n_q/2-1}\cos \left(\frac{V_{\mathrm{mq}}}{V_{\pi q}}\pi \right)& \left(\mathrm{Equation}8\right)\end{array}$

In addition, the present disclosure provides the optical modulation device, further including:

an I-component data signal voltage adjustment unit that adjusts, when the I-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time becomes equal to the intensity of the QAM signal at the drift occurrence time, the voltage of the I-component data signal such that the I-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time does not become equal to the intensity of the QAM signal at the drift occurrence time; and

a Q-component data signal voltage adjustment unit that adjusts, when the Q-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time becomes equal to the intensity of the QAM signal at the drift occurrence time, the voltage of the Q-component data signal such that the Q-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time does not become equal to the intensity of the QAM signal at the drift occurrence time.

According to this configuration, the present disclosure is applicable since it is possible to select whether to maximize or minimize an optical power after adjusting the voltages of I and Q-component data signals even when it is not possible to select whether to maximize or minimize the optical power in automatically controlling the bias voltage of an optical modulator.

In addition, the present disclosure provides the optical modulation device, further including:

an I-component dither voltage signal output unit that outputs a signal, which is superposed on the signal having the I-component bias voltage and has an I-component dither voltage having a frequency ω_d, to the I-component optical modulation unit; and

a Q-component dither voltage signal output unit that outputs a signal, which is superposed on the signal having the Q-component bias voltage and has a Q-component dither voltage having a phase difference π/2 between the frequency ω_d and the I-component dither voltage, to the Q-component optical modulation unit, wherein

the I-component bias voltage signal adjustment unit performs synchronous detection of the QAM signal with the signal having the I-component dither voltage and adjusts the I-component bias voltage such that a result of the synchronous detection becomes 0 to optimize the intensity of the QAM signal to an extreme value, and

the Q-component bias voltage signal adjustment unit performs synchronous detection of the QAM signal with the signal having the Q-component dither voltage and adjusts the Q-component bias voltage such that a result of the synchronous detection becomes 0 to optimize the intensity of the QAM signal to an extreme value.

According to this configuration, it is possible to maximize or minimize an optical power with the application of dithering to an I-component bias voltage and a Q-component bias voltage.

In addition, the present disclosure provides the optical modulation device, further including:

a phase shift unit bias voltage signal output unit that outputs a signal having a phase shift unit bias voltage, which is used to adjust the phase difference between the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to π/2, to the phase shift unit; and

a phase shift unit bias voltage signal adjustment unit that performs synchronous detection of the QAM signal with a signal having a frequency 2ω_d twice the frequency ω_d and adjusts the phase shift unit bias voltage such that a result of the synchronous detection becomes 0 to adjust the phase difference between the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to π/2.

According to this configuration, it is possible to adjust the phase difference between an I-component and a Q-component to π/2 with the application of dithering to an I-component bias voltage and a Q-component bias voltage.

In addition, the present disclosure provides the optical modulation device, further including:

an I-component bias voltage signal sweeping unit that sweeps the I-component bias voltage in a range of 2 V_πi or larger at an initializing operation of the optical modulation device when the half wavelength voltage of the I-component optical modulation unit is V_πi;

a Q-component bias voltage signal sweeping unit that sweeps the Q-component bias voltage in a range of 2 V_πq or larger at the initializing operation of the optical modulation device when the half wavelength voltage of the Q-component optical modulation unit is V_πq;

a phase shift unit bias voltage signal initialization unit that

• • performs synchronous detection of the QAM signal with the signal having frequency 2ω_d twice the frequency ω_d while causing the I-component bias voltage signal sweeping unit and the Q-component bias voltage signal sweeping unit to perform the sweeping operation at the initializing operation of the optical modulation device and
  • initializes the phase shift unit bias voltage such that a result of the synchronous detection becomes 0 or becomes the closest to 0 at all times in a range of the sweeping operation of I-component bias voltage signal sweeping unit and the Q-component bias voltage signal sweeping unit; and

a post-initialization adjustment unit that causes the I-component bias voltage signal adjustment unit, the Q-component bias voltage signal adjustment unit, and the phase shift unit bias voltage signal adjustment unit to perform adjustment after the initialization of the phase shift unit bias voltage by the phase shift unit bias voltage signal initialization unit.

According to this configuration, after setting a phase shift unit bias voltage at an optimum value or a value close to the optimum value in a post-initialization state in which an I-component bias voltage, a Q-component bias voltage, and the phase shift unit bias voltage deviate from optimum values, it is possible to set the I-component bias voltage, the Q-component bias voltage, and the phase shift unit bias voltage at final optimum values in an after-initialization state in which the phase shift unit bias voltage is set at the optimum value or the value close to the optimum value.

In addition, the present disclosure provides the optical modulation device, further including:

a controller unit that

• • stops the Q-component bias voltage signal adjustment unit when operating the I-component bias voltage signal adjustment unit and
  • stops the I-component bias voltage signal adjustment unit when operating the Q-component bias voltage signal adjustment unit.

According to this configuration, it is possible to perform prompt and reliable convergence to a correct equilibrium point with the control of an I-component bias voltage and a Q-component bias voltage in a time sharing manner.

In addition, the present disclosure provides the optical modulation device, wherein

a progression {V_mi} becomes an arithmetic progression when the voltage of the I-component data signal is V_mi (where m_i=0, 1, . . . , n_i/2−1 when the number of the I-component data signal level is n_i), and

a progression {V_mq} becomes an arithmetic progression when the voltage of the Q-component data signal is V_mq (where m_q=0, 1, . . . , n_q/2−1 when the number of the Q-component data signal level is n_q).

According to this configuration, it is possible to easily obtain a constellation in a lattice pattern.

In addition, the present disclosure provides the optical modulation device, wherein

a progression {sin(V_mi/V_πi×π/2)} becomes an arithmetic progression when the voltage of the I-component data signal is V_mi (where m_i=0, 1, . . . , n_i/2−1 when the number of the I-component data signal level is n_i) and the half wavelength voltage of the I-component optical modulation unit is V_πi, and

a progression {sin(V_mq/V_πq×π/2)} becomes an arithmetic progression when the voltage of the Q-component data signal is V_mq (where m_q=0, 1, . . . , n_q/2−1 when the number of the Q-component data signal level is n_q) and the half wavelength voltage of the Q-component optical modulation unit is V_πq.

According to this configuration, it is possible to obtain an ideal constellation in a lattice pattern.

ADVANTAGEOUS EFFECTS OF DISCLOSURE

According to the present disclosure, it is possible to provide a technique for appropriately selecting whether to maximize or minimize an optical power in automatically controlling the bias voltage of an optical modulator.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a diagram showing the configuration of an optical modulation device in the related art;

FIG. 2 is a diagram showing the characteristics of the optical modulation device in a case in which no bias drift occurs;

FIG. 3 is a diagram showing a constellation in a case in which no bias drift occurs;

FIG. 4 is a diagram showing the characteristics of the optical modulation device in a case in which a bias drift occurs;

FIG. 5 is a diagram showing a constellation in a case in which a bias drift occurs;

FIG. 6 is a diagram showing the relationship between the drift of a bias voltage and the intensity of an optical signal;

FIG. 7 is a diagram showing the relationship between the drift of the bias voltage and the intensity of the optical signal;

FIG. 8( a) is a diagram showing the characteristics of an optical modulation device according to the present disclosure;

FIG. 8( b) is a diagram showing the characteristics of an optical modulation device according to the present disclosure;

FIG. 8( c) is a diagram showing the characteristics of an optical modulation device according to the present disclosure;

FIG. 9 is a diagram showing the constellation of an optical 32 QAM signal;

FIG. 10 is a diagram showing the configuration of an optical modulation device according to a first embodiment;

FIG. 11 is a diagram showing the configuration of an optical modulation device according to a second embodiment;

FIG. 12 is a diagram showing the configuration of an optical modulation device according to a third embodiment;

FIG. 13 is a diagram showing the configuration of an optical modulation device according to a fourth embodiment;

FIG. 14 is a diagram showing the configuration of the optical modulation device according to the fourth embodiment; and

FIG. 15 is a flowchart showing processing at an initializing operation according to a fifth embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

A description will be given, with reference to the accompanying drawings, of embodiments of the present disclosure. The embodiments, which will be described below, are exemplified to carry out the present disclosure and do not intend to limit the present disclosure. Here, the constituents with the same symbols in the specification and drawings are the same each other.

(Outline of the Present Disclosure)

Attention will be given to a case in which a dual-type IQ optical modulator M generates an optical QAM signal whose order is the square of n. Here, it is assumed that n is a positive even number. Each of a differential signal (Data1- Data1) applied to a first driving electrode 61 and a differential signal (Data2- Data2) applied to a second driving electrode 71 has n differential voltages V₀, V₁, V₂, . . . , −V₂, −V₁, and −V₀. Here, it is assumed that V₀>V₁>V₂ > . . . >−V₂>−V₁>−V₀ is established. When the half wavelength voltages of the first and second optical modulation units 2 and 3 are V_π, the upper limit of |V₀| is desirably set at V_π to prevent the occurrence of unnecessary overshoot. However, in order to realize non-linear bias dithering described in Non-Patent Literature 1, the absolute value of V having a minimum absolute value may be less than V_π and the absolute value of V having another absolute value may be greater than or equal to V_π.

If all bias voltages have ideal values, the electric field of the output light of the first and second optical modulation units 2 and 3 is expressed by sin(±V_m/V_π×π/2) as shown in FIG. 2. Here, m is an integer of 0 to n/2−1. In addition, since optical power P_total is proportional to the square of the electric field of light, P_total is expressed by Equation 9. Hereinafter, a non-essential coefficient will be omitted.

$\begin{array}{cc}{P}_{\mathrm{total}}=4n\sum _{m=0}^{n/2-1}{\sin }^2\left(\frac{V_m}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}9\right)\end{array}$

A method for deriving Equation 9 will be described below. As shown in FIG. 3, the sum of the optical power is expressed by Equation 10 for the four symbols of the first quadrant of 16QAM.

$\begin{array}{cc}\begin{array}{c}{P}_{\mathrm{total}}=\left(E_{11}^2+E_{21}^2\right)+\left(E_{11}^2+E_{22}^2\right)+\left(E_{12}^2+E_{21}^2\right)+\\ \left(E_{12}^2+E_{22}^2\right)\\ =2\left(E_{11}^2+E_{12}^2+E_{21}^2+E_{22}^2\right)\\ =2\times 2\times \left\{{\sin }^2\left(\frac{V_0}{V_{\pi }}\frac{\pi }{2}\right)+{\sin }^2\left(\frac{V_1}{V_{\pi }}\frac{\pi }{2}\right)\right\}\end{array}& \left(\mathrm{Equation}10\right)\end{array}$

For the 16 symbols of all the quadrants of the 16QAM, the sum of the optical power is expressed by Equation 11.

$\begin{array}{cc}{P}_{\mathrm{total}}=4\times 2\times 2\times \left\{{\sin }^2\left(\frac{V_0}{V_{\pi }}\frac{\pi }{2}\right)+{\sin }^2\left(\frac{V_1}{V_{\pi }}\frac{\pi }{2}\right)\right\}& \left(\mathrm{Equation}11\right)\end{array}$

Further, for a general n, Equation 9 is established.

Here, for simplicity, it is assumed that only V_bias1 drifts and shifts by V_d. As shown in FIG. 5, some absolute values of V_m increase or decrease, and P_total changes from Equation 9 to Equation 12.

$\begin{array}{cc}{P}_{\mathrm{total}}=n\sum _{m=0}^{n/2-1}{\sin }^2\left(\frac{V_m+V_d}{V_{\pi }}\frac{\pi }{2}\right)+n\sum _{m=0}^{n/2-1}{\sin }^2\left(\frac{V_m-V_d}{V_{\pi }}\frac{\pi }{2}\right)+2n\sum _{m=0}^{n/2-1}{\sin }^2\left(\frac{V_m}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}12\right)\end{array}$

A method for deriving Equation 12 will be described below. Since V_bias1 shifts by V_d, Equation 13 is established.

$\begin{array}{cc}{E}_{11}=\sin \left(\frac{V_0+V_d}{V_{\pi }}\frac{\pi }{2}\right),E_{12}=\sin \left(\frac{V_1+V_d}{V_{\pi }}\frac{\pi }{2}\right),E_{13}=\sin \left(\frac{-V_1+V_d}{V_{\pi }}\frac{\pi }{2}\right)=-\sin \left(\frac{V_1-V_d}{V_{\pi }}\frac{\pi }{2}\right),E_{14}=\sin \left(\frac{-V_0+V_d}{V_{\pi }}\frac{\pi }{2}\right)=-\sin \left(\frac{V_0-V_d}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}13\right)\end{array}$

Since V_bias2 does not shift, Equation 14 is established.

$\begin{array}{cc}{E}_{21}=\sin \left(\frac{V_0}{V_{\pi }}\frac{\pi }{2}\right),E_{22}=\sin \left(\frac{V_1}{V_{\pi }}\frac{\pi }{2}\right),E_{23}=\sin \left(\frac{-V_1}{V_{\pi }}\frac{\pi }{2}\right)=-\sin \left(\frac{V_1}{V_{\pi }}\frac{\pi }{2}\right),E_{24}=\sin \left(\frac{-V_0}{V_{\pi }}\frac{\pi }{2}\right)=-\sin \left(\frac{V_0}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}14\right)\end{array}$

When the sum of the square of each electric field is calculated, Equation 12 is established for a general n.

When P_total is differentiated by V_d, Equation 15 is obtained.

$\begin{array}{cc}\frac{d{P}_{\mathrm{total}}}{d{V}_d}=2n\sin \left(\frac{V_d}{V_{\pi }}\pi \right)\sum _{m=0}^{n/2-1}\cos \left(\frac{V_m}{V_{\pi }}\pi \right)& \left(\mathrm{Equation}15\right)\end{array}$

Because of the periodicity of an MZI type optical modulator, ideal V_bias1 is obtained not only under V_d=0 but also under V_d=±2 V_π, ±4 V_π, etc. As is evident from Equation 15, the optical power P_total shows a maximum or minimum value under V_d=0, ±2 V_π, ±4 V_π, etc. In the following description, the vicinity of V_d=0 will be considered.

FIGS. 8( a), (b) and (c) show the characteristics of an optical modulation device according to the present disclosure. If it is not certain whether a value shown in Equation 16 is a positive or negative value, it is not certain whether the optical power P_total shows a maximum or minimum value under V_d=0. In view of this, determination is made depending on the sign of the value shown in Equation 16.

$\begin{array}{cc}\sum _{m=0}^{n/2-1}\cos \left(\frac{V_m}{V_{\pi }}\pi \right)& \left(\mathrm{Equation}16\right)\end{array}$

Since dP_total/dV_d shown in Equation 15 becomes 0 under V_d=0 and increases with an increase in V_d if the value shown in Equation 16 is positive, the optical power P_total has a minimum value under V_d=0. Accordingly, when V_bias1 is controlled such that P_total becomes minimum, an ideal bias voltage is obtained.

Since dP_total/dV_d shown in Equation 15 becomes 0 under V_d=0 and decreases with an increase in V_d if the value shown in Equation 16 is negative, the optical power P_total has a maximum extreme value under V_d=0. Accordingly, when V_bias1 is controlled such that P_total becomes maximum, an ideal bias voltage is obtained.

Since dP_total/dV_d shown in Equation 15 becomes 0 for any value of V_d if the value shown in Equation 16 is 0, the optical power P_total becomes equal for any value of V_d. Accordingly, when V_bias1 is controlled such P_total becomes maximum or minimum, it is not certain whether an ideal bias voltage is obtained. In view of this, driving signals V₀, V₁, V₂, . . . , −V₂, −V₁, and −V₀, with which the value shown in Equation 16 becomes 0, are prevented. Meanwhile, if the optical modulator is designed such that the number of level n and the amplitude V_m of a driving signal are constant at all times during the operation period of the optical modulation device, the value shown in Equation 16 becomes also constant. In this case, determination as to whether P_total should be maximum or minimum is uniquely made when the optical modulation device is designed. In such a case, there is no need to repeat determination using Equation 16 as occasion demands during the operation period.

The dual drive type IQ optical modulator is used in the above description, however a single drive type IQ optical modulator applies single end drive signals instead of differential driving signals to the first and second driving electrodes 61 and 71. If the potentials of such driving signals with respect to the ground are set at V₀, V₁, V₂, . . . , −V₂, −V₁, and −V₀, completely the same discussion as the above is established.

The above description refers to an example in which a constellation is a lattice and square array. However, in optical electric field generated by other type of QAM signal, some of the lattice and square array is cut out, in order to facilitate the multiplexing of a binary signal.

As a specific example, FIG. 9 shows the constellation of a 32 QAM signal. In the 32 QAM signal, the four apexes of the constellation of a 36 QAM signal are prohibited, and only 32 types of optical electric fields are used. With the use of the 32 types of optical electric fields, it is possible to efficiently multiplex the five binary signals, because 2⁵=32.

For the 32 QAM signal, if a term corresponding to a prohibited rule is divided from a series and Equation 9 is deformed into Equation 17 to calculate P_total in Equation 9, the same discussion as the above description is established. The same applies to QAM signals with other irregular constellations.

$\begin{array}{cc}{P}_{\mathrm{total}}=4n\sum _{m=0}^{n/2-1}{\sin }^2\left(\frac{V_m}{V_{\pi }}\frac{\pi }{2}\right)-4{\sin }^2\left(\frac{V_0}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}17\right)\end{array}$

First Embodiment

FIG. 10 shows the configuration of an optical modulation device according to a first embodiment. In the dual drive type IQ optical modulator M, the differential drive signals Data1 and Data1 and Data2 and Data2 output from the first and second driving amplifiers 6 and 7 are applied to the first and second driving electrodes 61 and 71. The first and second quaternary data signals are input to the first and second driving amplifiers 6 and 7 and the output amplitudes of the signals are adjusted by the first and second amplitude adjustment units 62 and 72. In the first embodiment, pre-emphasis is not used, and the four signal levels V₀, V₁, −V₁, and −V₀ of the differential signals added to the first and second driving electrodes 61 and 71 are set at even intervals. In other words, (Data1- Data1) and (Data2- Data2) have the four values V₀, V₁, −V₁, and −V₀, and |V₀−V₁|=|V₁−(−V₁)| is established.

As a first specific example, it is assumed that the interval between the four levels of driving signals is set at 0.6 V_π. In this case, V₀=0.9 V_π and V₁=0.3 V_π are established. When these values are substituted into Equation 16, −0.36 is obtained. Based on the above discussion, when V_bias1 and V_bias2 are controlled such that P_total becomes maximum on this occasion, the most suitable modulation state is obtained.

As a second specific example, it is assumed that the interval between the four levels of driving signals is set at 0.4 V_π. In this case, V₀=0.6 V_π and V₁=0.2 V_π are established. When these values are substituted into Equation 16, 0.5 is obtained. Based on the above discussion, when V_{bias 1} and V_bias2 are controlled such that P_total becomes minimum on this occasion, the most suitable modulation state is obtained.

As a third specific example, it is assumed that the interval between the four levels of driving signals is set at 0.5 V_π. In this case, V₀=0.75 V_π and V₁=0.25 V_π are established. When these values are substituted into Equation 16, 0 is obtained. Based on the above discussion, such driving signals are prohibited in the present disclosure. If the value of Equation 16 becomes 0, the values V₀, V₁, −V₁, and −V₀ are changed to other values using the first and second amplitude adjustment units 62 and 72.

In the IQ optical modulator M, the data bias voltages V_bias1 and V_bias2 output from the first and second bias power supplies 8 and 9 are applied to the first and second bias electrodes 81 and 91, respectively. Each of the first and second optical modulation units 2 and 3 has two arms, however the remaining one arm of each of the first and second optical modulation units 2 and 3 to which no data bias voltage is applied is grounded.

Dithering is applied to V_bias1 and V_bias2 by a first oscillator 82, a first adder 83, a second oscillator 92, and a second adder 93 at a frequency ω_d. The amplitude of the dithering is kept at a level at which the quality of a 16 QAM signal is not degraded. The frequency ω_d is set to be substantially lower than the baud rate of the 16 QAM signal and be in the order of kHz at most. Considering that the phase difference between an I-component and a Q-component is π/2, the phases of the first and second oscillators 82 and 92 are set to be orthogonal to each other. In the first embodiment, it is assumed that the output of the first oscillator 82 is expressed as cos(ω_dt) and that of the second oscillator 92 is expressed as sin(ω_dt). Here, t expresses the time. Accordingly, dithering synchronized with cos(ω_dt) is applied to V_bias1, and dithering synchronized with sin(ω_dt) is applied to V_bias2.

The 16 QAM signal output from the IQ optical modulator M is branched by an optical demultiplexing coupler 11 and input to an optical power monitor 12. The band of the optical power monitor 12 may be nearly twice the frequency ω_d of the dithering. Some of the output of the optical power monitor 12 is input to the first and second synchronous detection circuits 84 and 94.

Here, it is assumed that a bias drift occurs in V_bias1 and V_bias1 shifts from its optimum value to a positive or negative side. The optical power of the 16 QAM signal becomes no longer an extreme value and increases or decreases depending on the symbol of Equation 16. Since the dithering of cos(ω_dt) is applied to V_bias1, it is possible to determine the positive and negative of the shift of the optical power if synchronous detection is performed by the first synchronous detection circuit 84 with the output of the first oscillator 82 as a reference clock. Based on the determination result, a control signal is sent to the first bias power supply 8 to give feedback to V_bias1 such that the synchronous detection result becomes 0. The state in which the synchronous detection result becomes 0 is realized by two states in which the optical power has a maximum or minimum extreme value, however the selection is uniquely determined by the positive and negative of Equation 16 and a method for applying a bias. As is the case with the control of V_bias1, the control of V_bias2 is performed using the second oscillator 92 and the second synchronous detection circuit 94.

V_bias3 output from a third bias power supply 10 is applied to an orthogonal bias electrode 101 of the IQ optical modulator M. No dithering is applied to V_bias3.

Here, it is assumed that a bias drift occurs in V_bias3 and V_bias3 shifts from its optimum value to a positive or negative side. At this time, as described in detail in Non-Patent Literature 1, the optical power of the 16 QAM signal has an extreme value at phases π/4 and 5π/4 for the frequency ω_d, so that the intensity modulation component of a frequency 2ω_d appears in the optical power of the 16 QAM signal. The phase of the intensity modulation component of the frequency 2ω_d is determined by the positive and negative of the shift of V_bias3.

Some of the output of the optical power monitor 12 is input to a third synchronous detection circuit 103. The third synchronous detection circuit 103 detects the intensity modulation component of the frequency 2ω_d from the optical power of the 16 QAM signal with the oscillation signal of the frequency 2ω_d output from the third oscillator 102 as a reference clock, determines the positive and negative of the shift of V_bias3 from the phase, and sends a control signal to the third bias power supply 10 to give a feedback such that V_bias3 becomes optimum.

In the first embodiment, the first, second, and third synchronous detection circuits 84, 94, and 103 are used. Here, these circuits may be integrated into one circuit to change the frequency of a reference clock between ω_d and 2ω_d in the one synchronous detection circuit. Or the frequency of the reference clock may be fixed at ω_d to detect a secondary higher harmonic wave having the frequency 2ω_d.

Second Embodiment

FIG. 11 shows the configuration of an optical modulation device according to a second embodiment. As the difference between the first and second embodiments, pre-emphasis is used to set the values of the electric fields E₁₁, E₁₂, E₁₃, and E₁₄ of the four types of light output from the first optical modulation unit 2 at even intervals and set the values of the electric fields E₂₁, E₂₂, E₂₃, and E₂₄ of the four types of light output from the second optical modulation unit 3 at even intervals. In other words, Equation 18 is established.E₁₁=E₂₁=sin(V₀/V_π×π/2)E₁₂=E₂₂=sin(V₁/V_π×π/2)E₁₃=E₂₃=sin(−V₁/V_π×π/2)E₁₄=E₂₄=sin(−V₀/V_π×π/2)|sin(V₀/V_π×π/2)−sin(V₁/V_π×π/2)|=2|sin(V₁/V_π×π/2)|  (Equation 18)

Required V₀ and V₁ are obtained when the first and second quaternary data signals are input to the first and second pre-emphasis units 63 and 73, respectively. The control of V_bias1 to V_bias3 is performed in the same manner as the first embodiment.

As a first specific example, it is assumed that E₁₁−E₁₂=E₁₂−E₁₃=E₁₃−E₁₄=E₂₁−E₂₂=E₂₂−E₂₃=E₂₃−E₂₄=0.63 is set. In this case, V₀=0.80 V_π and V₁=0.21 V_π are established. When these values are substituted into Equation 16, −0.006 is obtained. Based on the above discussion, when V_bias1 and V_bias2 are controlled such that P_total becomes maximum on this occasion, the most suitable modulation state is obtained.

As a second specific example, it is assumed that E₁₁−E₁₂=E₁₂−E₁₃=E₁₃−E₁₄=E₂₁−E₂₂=E₂₂−E₂₃=E₂₃−E₂₄=0.60 is set. In this case, V₀=0.71 V_π and V₁=0.19 V_π are established. When these values are substituted into Equation 16, 0.20 is obtained. Based on the above discussion, when V_bias1 and V_bias2 are controlled such that P_total becomes minimum on this occasion, the most suitable modulation state is obtained.

Third Embodiment

FIG. 12 shows the configuration of an optical modulation device according to a third embodiment. The third embodiment is different from the first embodiment in two points.

A first different point is that this embodiment uses a single driving type IQ optical modulator M instead of a dual driving type IQ optical modulator M. Each of the first and second driving amplifiers 6 and 7 performs a single-phase output, and their outputs Data1 and Data2 have the four signal levels V₀, V₁, −V₁, and −V₀ relative to the ground level. Although twice the output amplitude of the first embodiment is required, there is an advantage in that the mounting area of the IQ optical modulator M can be reduced.

A second different point is that the first and second bias power supplies 8 and 9 perform a single-phase output, their DC voltages are converted into differential signals by the first and second dithering amplifiers 85 and 95 after having the outputs of the first and second oscillators 82 and 92 superposed thereon with the first and second adders 83 and 93, and applied to the first and second bias electrodes 81 and 91. Accordingly, in FIG. 1, dithering synchronized with cos(ω_dt) and cos(ω_dt+π) is applied to V_bias1 and V′_bias1, and dithering synchronized with sin(ω_dt) and sin(ω_dt+π) is applied to V_bias2 and V′_bias2.

Fourth Embodiment

FIGS. 13 and 14 show the configuration of an optical modulation device according to a fourth embodiment. The fourth embodiment is different from the first embodiment in two points.

A first different point is that the first oscillator 82, the second oscillator 92, the third oscillator 102, the first synchronous detection circuit 84, the second synchronous detection circuit 94, and the third synchronous detection circuit 103 are controlled by a controller 13 and operated in a time sharing manner for each bias voltage.

In this embodiment, first, only the first oscillator 82 and the first synchronous detection circuit 84 are driven to optimize V_bias1. The control of V_bias1 is stopped when V_bias1 is stabilized. Next, only the second oscillator 92 and the second synchronous detection circuit 94 are driven to optimize the control of V_bias2. The control of V_bias2 is stopped when V_bias2 is stabilized. Then, only the third oscillator 102 and the third synchronous detection circuit 103 are driven to optimize the control of V_bias3. The control of V_bias3 is stopped when V_bias3 is stabilized, and the process returns to the control of V_bias1 again. With the sequential bias control in a time sharing manner as described above, it becomes possible to perform more prompt and reliable convergence to an equilibrium point.

When the third oscillator 102 and the third synchronous detection circuit 103 are driven, the first oscillator 82 and the second oscillator 92 are further driven and the control of V_bias3 is optimized by the dithering of V_bias1 and V_bias2 as described in detail in Non-Patent Literature 1.

A second different point is that an optical multiplexing/demultiplexing coupler 14, a photo detector 15, and an optical power monitor terminal 151 internal to the IQ optical modulator M are used instead of a second optical coupler 5, the optical demultiplexing coupler 11, and the optical power monitor 12 external to the IQ optical modulator M. Here, there is a direction for using the commercially-available IQ optical modulator M shown in FIG. 14.

The optical multiplexing/demultiplexing coupler 14 is used instead of the second optical coupler 5, and the multiplexed output light of the first and second optical modulation units 2 and 3 is demultiplexed into two. Among the demultiplexed light, one is output from the IQ optical demodulator M while the other is input to the photo detector 15. The output light of the first photo modulation unit 2 and that of the second photo modulation unit 3 interfere with each other inside the optical multiplexing/demultiplexing coupler 14, however the interference intensities of the output light of the IQ optical modulator M and the input light to the photo detector 15 are inverted due to the asymmetry of an optical waveguide coupler. In other words, the input light to the photo detector 15 weakens when the output light of the IQ optical modulator M intensifies due to the interference, while the input light to the photo detector 15 intensifies when the output light of the IQ optical modulator M weakens due to the interference. For this reason, this embodiment requires the inversion of the positive and negative of the control signal compared with the first to third embodiments.

Fifth Embodiment

The above description mainly refers to a case in which two of the three types of bias voltages V_bias1 to V_bias3 are kept at ideal voltages and the rest bias voltage starts drifting. For example, in FIG. 5, Equation 15 is derived in a state in which a drift occurs in only V_bias1 and the other two types of bias voltages are set at optimum values.

However, since it is hard to immediately set the three types of bias voltages V_bias1 to V_bias3 at optimum values at the starting operation of a commercial transmitter, there is a likelihood of all the three types of bias voltages greatly deviating from ideal values.

A commercial transmitter is required to have a bias voltage control algorithm by which all the three types of bias voltages are accurately converged into optimum values even if the transmitter starts from such an initial state. A fifth embodiment refers to such an algorithm. For simplicity, the fifth embodiment exemplifies the control of a transmitter that generates an optical QAM signal in which an order is the square of an even number n and a constellation has an n×n lattice pattern.

First, a variable is defined as follows. It is assumed that in a state in which dithering is not applied immediately after the start of the transmitter, the difference between the value of V_bias1 and an optimum value is V_d1 and the difference between the value of V_bias2 and an optimum value is V_d2. Since V_bias3 is not also limited to an optimum value immediately after the start of the transmitter, θ₃ is not limited to ±π/2 in the following Equation.

It is assumed that the voltages of dithering signals superposed on V_bias1 and V_bias2 are expressed as A_d cos(ω_dt) and A_d sin(ω_dt), respectively. Here, A_d expresses a dither amplitude, and t expresses the time. The optical power P_total of the QAM signal of an order n² slightly fluctuates depending on ω_dt as described above, however the optical power under ω_dt=φ is expressed here as P_total(φ) in order to correctly express the slight fluctuation.

When Equation 12 is rewritten from a case in which only V_bias1 deviates from an ideal value to a case in which all the three types of bias voltages deviate from ideal values, Equation 19 is obtained.

$\begin{array}{cc}{P}_{\mathrm{total}}\left(\phi \right)=n\sum _{L=0}^{n/2-1}{\sin }^2\left(\frac{V_L+{\delta }_1\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)+n\sum _{L=0}^{n/2-1}{\sin }^2\left(\frac{V_L-{\delta }_1\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)+n\sum _{M=0}^{n/2-1}{\sin }^2\left(\frac{V_M+{\delta }_2\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)+n\sum _{M=0}^{n/2-1}{\sin }^2\left(\frac{V_M-{\delta }_2\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)+2\cos {\theta }_3\sum _{L=0}^{n/2-1}\sum _{M=0}^{n/2-1}[\begin{array}{c}\left\{\sin \left(\frac{V_L+{\delta }_1\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)+\sin \left(\frac{-V_L+{\delta }_1\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)\right\}\times \\ \left\{\sin \left(\frac{V_M+{\delta }_2\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)+\sin \left(\frac{-V_M+{\delta }_2\left(\phi \right)}{V_{\pi }}\frac{\pi }{2}\right)\right\}\end{array}]& \left(\mathrm{Equation}19\right)\end{array}$

However, δ₁(φ) and δ₂(φ) are expressed by Equations 20 and 21, respectively.δ₁(φ)=V_d1+A_d cos φ  (Equation 20)δ₂(φ)=V_d2+A_d sin φ  (Equation 21)

When P_total(π)−P_total(0) is calculated to obtain the approximate value of the output of the first synchronous detection circuit 84 described in the first embodiment, Equation 22 is obtained.

$\begin{array}{cc}{P}_{\mathrm{total}}\left(\pi \right)-P_{\mathrm{total}}\left(0\right)=-2n\sin \left(\frac{A_d}{V_{\pi }}\pi \right)\sin \left(\frac{V_{d1}}{V_{\pi }}\pi \right)\sum _{L=0}^{n/2-1}\cos \left(\frac{V_L}{V_{\pi }}\pi \right)-16\cos {\theta }_3\sin \left(\frac{A_d}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_{d1}}{V_{\pi }}\frac{\pi }{2}\right)\sin \left(\frac{V_{d2}}{V_{\pi }}\frac{\pi }{2}\right)\times \sum _{L=0}^{n/2-1}\sum _{M=0}^{n/2-1}\cos \left(\frac{V_L}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_M}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}22\right)\end{array}$

In addition, when P_total(3π/2)−P_total(π/2) is calculated to obtain the approximate value of the output of the second synchronous detection circuit 94 described in the first embodiment, Equation 23 is obtained.

$\begin{array}{cc}{P}_{\mathrm{total}}\left(3\pi /2\right)-P_{\mathrm{total}}\left(\pi /2\right)=-2n\sin \left(\frac{A_d}{V_{\pi }}\pi \right)\sin \left(\frac{V_{d2}}{V_{\pi }}\pi \right)\sum _{M=0}^{n/2-1}\cos \left(\frac{V_M}{V_{\pi }}\pi \right)-16\cos {\theta }_3\sin \left(\frac{A_d}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_{d2}}{V_{\pi }}\frac{\pi }{2}\right)\sin \left(\frac{V_{d1}}{V_{\pi }}\frac{\pi }{2}\right)\times \sum _{L=0}^{n/2-1}\sum _{M=0}^{n/2-1}\cos \left(\frac{V_L}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_M}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}23\right)\end{array}$

If all of V_bias1 to V_bias3 are optimum, θ₃=±π/2 and V_d1=V_d2=0 are established and thus values obtained from Equations 22 and 23 are 0. If only V_bias3 is optimum and V_bias1 and V_bias2 are inappropriate, the values obtained from Equations 22 and 23 depend on V_d1 or V_d2 and the positive and negative of Equations 22 and 23 is determined by the positive and negative of Equation 16. This result is a mathematical expression showing that optimum V_bias1 and V_bias2 can be achieved if V_bias1 and V_bias2 are changed such that the synchronous detection results of the first and second synchronous detection circuits 84 and 94 become 0 in the same manner as the feedback control of V_bias1 and V_bias2 in the first embodiment.

However, since the second terms on the right side of Equations 22 and 23 have any value other than 0 if all of V_bias1 to V_bias3 are inappropriate at the stating operation of the transmitter, V_bias1 and V_bias2 cannot be promptly converged into optimum values.

Here, all of V_bias1 to V_bias3 finally reach optimum values if V_bias1 to V_bias3 are successively corrected by the alternate use of the two circuits, i.e. the first and second synchronous detection circuits 84 and 94 that perform the feedback control of V_bias1 and V_bias2 in the first embodiment, and the one circuit, i.e. the third synchronous detection circuit 103 that performs the feedback control of V_bias3, however it takes time to converge V_bias1 to V_bias3 into optimum values depending on their initial state.

In order to prevent this problem, it is required to first adjust V_bias3 to an optimum value and then adjust V_bias1 and V_bias2 to optimum values at the starting operation of the transmitter or it is required to first adjust V_bias3 to a value close to an optimum value and then adjust V_bias3 to a final optimum value while adjusting V_bias1 and V_bias2 to optimum values.

When P_total(7π/4)+P_total(3π/4)−P_total(5π/4)−P_total(π/4) is calculated to obtain the approximate value of the output of the third synchronous detection circuit 103, Equation 24 is obtained.

$\begin{array}{cc}{P}_{\mathrm{total}}\left(7\pi /4\right)+P_{\mathrm{total}}\left(3\pi /4\right)-P_{\mathrm{total}}\left(5\pi /4\right)-P_{\mathrm{total}}\left(\pi /4\right)=-32\cos {\theta }_3{\sin }^2\left(\frac{A_d/\sqrt{2}}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_{d1}}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_{d2}}{V_{\pi }}\frac{\pi }{2}\right)\times \sum _{L=0}^{n/2-1}\sum _{M=0}^{n/2-1}\cos \left(\frac{V_L}{V_{\pi }}\frac{\pi }{2}\right)\cos \left(\frac{V_M}{V_{\pi }}\frac{\pi }{2}\right)& \left(\mathrm{Equation}24\right)\end{array}$

From Equation 24, the following three conclusions (1) to (3) can be obtained.

(1) If V_bias1 and V_bias2 are values close to optimum values and V_d1 and V_d2 are close to 0, the value obtained from Equation 24 becomes 0 on the condition that V_bias3 is an optimum value and θ₃=±π/2 is satisfied. This result is a mathematical expression showing that optimum V_bias3 can be achieved if V_bias3 is changed such that the synchronous detection result of the third synchronous detection circuit 103 becomes 0 in the same manner as the feedback of V_bias3 in the first embodiment.

(2) However, in a case in which V_bias1 and V_bias2 greatly deviate from optimum values and V_d1=±V_π or V_d2=±V_π is satisfied, the value obtained from Equation 24 becomes 0 even if V_bias3 is not an optimum value and θ₃=±π/2 is not satisfied.

(3) In a case in which V_d1 and V_d2 are forcibly swept in the range of +V_π to −V_π or in a greater range, the value obtained from Equation 24 during the forcible sweeping becomes 0 at all times on the condition that V_bias3 is an optimum value and θ₃=±π/2 is satisfied.

Since a commercial signal is not sent and received at the starting operation of the commercial transmitter, there is no effect on services even if V_bias1 and V_bias2 are forcibly swept.

By repeatedly changing V_bias1 and V_bias2 in the range of +V_π to −V_π for several V_bias3 with V_bias3 kept at a constant value and searching for V_bias3 such that the synchronous detection result of the third synchronous detection circuit 103 becomes 0 or nearly 0 at all times during the sweeping period, it is possible to set V_bias3 at an optimum value or a value close to the optimum value.

If V_bias3 can be set at an optimum value or a value close to the optimum value, it is possible to promptly converge V_bias1 and V_bias2 into optimum values using the first and second synchronous detection circuits 84 and 94 in the same manner as the first to fourth embodiments.

Hereinafter, a description will be given, with reference to the flowchart of FIG. 15, of processing at the initializing operation described above. This flow can be realized using, for example, a Field-Programmable Gate Array (FPGA).

First, the first bias power supply 8, the second bias power supply 9, and the third bias power supply 10 are placed in a forcible sweeping mode (step S1). Although these bias power supplies are controlled by the feedback signals output from the first, second, and third synchronous detection circuits 84, 94, and 103 in the feedback mode in the first to forth embodiments, they are set by a command output from a control circuit such as a FPGA in the forcible sweeping mode in the fifth embodiment.

Next, the notch widths ΔV₁, ΔV₂, ΔV₃ of bias values are set (step S2). A bearing force against the circuit noises of the synchronous detection circuits becomes greater if these values are smaller, however it requires a long time to perform a starting operation if these values are too small. Each of ΔV₁ to ΔV₃ is set at a value smaller than twice the half wavelength voltages V_π of V_bias1 to V_bias3.

Then, an array variable DATA(i) and an array variable BIAS3(i) are cleared to 0 (step S3), and the first and second oscillators 82 and 92 are oscillated (step S4). However, (i) expresses an array element. In addition, the unit of the array variable BIAS3(i) may be expressed not only by a voltage but also by an internal expression form used in the device. Before the start of a first loop, 0 is substituted into an integer type variable i and V_bias3 are set at 0 V (step S5).

In the first loop, first, V_bias3 is substituted into BIAS3(i) (step S6). Inside the first loop, a second loop is nested. Before the start of the second loop, V_bias1 and V_bias2 are set at 0 V (step S7).

In the second loop, first, the integer type variable i and V_bias3 are kept constant, the absolute values of the output of the third synchronous detection circuit 103 are checked one by one (step S8) while V_bias1 and V_bias2 are simultaneously increased at the notch widths ΔV₁ and ΔV₂ (step S10), and the largest one of the absolute values thus obtained is recorded on the DATA(i) (step S9). When V_bias1 and V_bias2 exceed 2 V_π, the procedure is released from the second loop (step S11).

When the second loop is released from the second loop, the integer type variable i is increased by 1 and V_bias3 is increased by the notch width ΔV₃ (step S12) to close the first loop. When V_bias3 exceeds 2 V_π, the procedure is released from the first loop (step S13).

i types of synchronous detection results are substituted into DATA(0) to DATA(i−1) when the first loop is ended, however a synchronous detection result with the smallest value is searched (step S14). If DATA(j) is the one with the smallest value, the voltage value expressed by BIAS3(j) is the closest to the optimum value of V_bias3 or the optimum value in a swept range and thus V_bias3 is set at the voltage value indicated by BIAS3(j) to end the initializing operation (step S15).

After the end of the initializing operation, V_bias1 and V_bias2 are optimized by the configurations described in the first to fourth embodiments to end the optimizing operation (step S16).

Effect of the Present Disclosure

The optical modulation device and the bias voltage control method according to the present disclosure can be effectively applied to a case in which a multi-QAM signal is sent, particularly to a case in which a quaternary or larger value QAM signal is sent.

REFERENCE SIGNS LIST

• M: IQ optical modulator
• 1: First optical coupler
• 2: First optical modulation unit
• 3: Second optical modulation unit
• 4: Optical phase shifter
• 5: Second optical coupler
• 6: First driving amplifier
• 7: Second driving amplifier
• 8: First bias power supply
• 9: Second bias power supply
• 10: Third bias power supply
• 11: Optical demultiplexing coupler
• 12: Optical power monitor
• 13: Controller
• 14: Optical multiplexing/demultiplexing coupler
• 15: Photo detector
• 61: First driving electrode
• 62: First amplitude adjustment unit
• 63: First pre-emphasis unit
• 71: Second driving electrode
• 72: Second amplitude adjustment unit
• 73: Second pre-emphasis unit
• 81: First bias electrode
• 82: First oscillator
• 83: First adder
• 84: First synchronous detection circuit
• 85: First dithering amplifier
• 91: Second bias electrode
• 92: Second oscillator
• 93: Second adder
• 94: Second synchronous detection circuit
• 95: Second dithering amplifier
• 101: Orthogonal bias electrode
• 102: Third oscillator
• 103: Third synchronous detection circuit
• 151: Optical power monitor terminal

Claims

1. An optical modulation device, comprising: an I-component optical modulation unit that modulates a continuous optical signal for an I-component of a QAM signal;a Q-component optical modulation unit that modulates a continuous optical signal for a Q-component of the QAM signal;a phase shift unit that shifts a phase of an optical signal on an input side and/or an output side for the I-component optical modulation unit and/or the Q-component optical modulation unit to adjust a phase difference between an output of the I-component optical modulation unit and an output of the Q-component optical modulation unit after the phase shift to π/2;a QAM signal generation unit that multiplexes the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to generate the QAM signal;an I-component data signal output unit that outputs an I-component data signal to the I-component optical modulation unit to cause the I-component optical modulation unit to modulate a continuous optical signal using the I-component data signal;a Q-component data signal output unit that outputs a Q-component data signal to the Q-component optical modulation unit to cause the Q-component optical modulation unit to modulate a continuous optical signal using the Q-component data signal;an I-component bias voltage signal output unit that outputs a signal having an I-component bias voltage corresponding to a null point of the I-component optical modulation unit to the I-component optical modulation unit;a Q-component bias voltage signal output unit that outputs a signal having a Q-component bias voltage corresponding to a null point of the Q-component optical modulation unit to the Q-component optical modulation unit;an I-component bias voltage signal adjustment unit that: determines whether an intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the I-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the I-component bias voltage,adjusts the I-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, andadjusts the I-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non- occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time; anda Q-component bias voltage signal adjustment unit that: determines whether the intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the Q-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the Q-component bias voltage,adjusts the Q-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, andadjusts the Q-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non- occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time.

2. The optical modulation device according to claim 1, further comprising: an I-component calculation unit that calculates a value shown in Equation 25 when a voltage of the I-component data signal is Vmi (where mi=0, 1, . . . , ni/2−1 when a number of the I-component data signal level is ni) and a half wavelength voltage of the I-component optical modulation unit is Vπi;

3. The optical modulation device according to claim 1, further comprising: an 1-component data signal voltage adjustment unit that adjusts, when the I-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time becomes equal to the intensity of the QAM signal at the drift occurrence time, the voltage of the I-component data signal such that the I-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time does not become equal to the intensity of the QAM signal at the drift occurrence time; anda Q-component data signal voltage adjustment unit that adjusts, when the Q-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time becomes equal to the intensity of the QAM signal at the drift occurrence time, the voltage of the Q-component data signal such that the Q-component bias voltage signal adjustment unit determines that the intensity of the QAM signal at the drift non-occurrence time does not become equal to the intensity of the QAM signal at the drift occurrence time.

4. The optical modulation device according to claim 1, further comprising: an I-component dither voltage signal output unit that outputs a signal, which is superposed on the signal having the I-component bias voltage and has an I-component dither voltage having a frequency ωd, to the I-component optical modulation unit; anda Q-component dither voltage signal output unit that outputs a signal, which is superposed on the signal having the Q-component bias voltage and has a Q-component dither voltage having a phase difference π/2 between the frequency ωd and the I-component dither voltage, to the Q-component optical modulation unit, whereinthe I-component bias voltage signal adjustment unit performs synchronous detection of the QAM signal with the signal having the I-component dither voltage and adjusts the I-component bias voltage such that a result of the synchronous detection becomes 0 to optimize the intensity of the QAM signal to an extreme value, andthe Q-component bias voltage signal adjustment unit performs synchronous detection of the QAM signal with the signal having the Q-component dither voltage and adjusts the Q-component bias voltage such that a result of the synchronous detection becomes 0 to optimize the intensity of the QAM signal to an extreme value.

5. The optical modulation device according to claim 4, further comprising: a phase shift unit bias voltage signal output unit that outputs a signal having a phase shift unit bias voltage, which is used to adjust the phase difference between the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to π/2, to the phase shift unit; anda phase shift unit bias voltage signal adjustment unit that performs synchronous detection of the QAM signal with a signal having a frequency 2ωd twice the frequency ωd and adjusts the phase shift unit bias voltage such that a result of the synchronous detection becomes 0 to adjust the phase difference between the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to π/2.

6. The optical modulation device according to claim 5, further comprising: an I-component bias voltage signal sweeping unit that sweeps the I-component bias voltage in a range of 2 Vπi or larger at an initializing operation of the optical modulation device when the half wavelength voltage of the I-component optical modulation unit is Vπi;a Q-component bias voltage signal sweeping unit that sweeps the Q-component bias voltage in a range of 2 Vπq or larger at the initializing operation of the optical modulation device when the half wavelength voltage of the Q-component optical modulation unit is Vπq;a phase shift unit bias voltage signal initialization unit that performs synchronous detection of the QAM signal with the signal having frequency 2ωd twice the frequency ωd while causing the I-component bias voltage signal sweeping unit and the Q-component bias voltage signal sweeping unit to perform the sweeping operation at the initializing operation of the optical modulation device, andinitializes the phase shift unit bias voltage such that a result of the synchronous detection becomes 0 or becomes the closest to 0 at all times in a range of the sweeping operation of I-component bias voltage signal sweeping unit and the Q-component bias voltage signal sweeping unit; anda post-initialization adjustment unit that causes the I-component bias voltage signal adjustment unit, the Q-component bias voltage signal adjustment unit, and the phase shift unit bias voltage signal adjustment unit to perform adjustment after the initialization of the phase shift unit bias voltage by the phase shift unit bias voltage signal initialization unit.

7. The optical modulation device according to claim 1, further comprising: a controller unit that stops the Q-component bias voltage signal adjustment unit when operating the I-component bias voltage signal adjustment unit, andstops the I-component bias voltage signal adjustment unit when operating the Q-component bias voltage signal adjustment unit.

8. The optical modulation device according to claim 1, wherein: a progression {Vmi} becomes an arithmetic progression when the voltage of the I-component data signal is Vmi (where mi=0, 1, . . . , ni/2−1 when the number of the I-component data signal level is ni), anda progression {Vmq} becomes an arithmetic progression when the voltage of the Q-component data signal is Vmq (where mq=0, 1, . . . , nq/2−1 when the number of the Q-component data signal level is nq).

9. The optical modulation device according to claim 1, wherein: a progression {sin(Vmi/Vπi×π/2)} becomes an arithmetic progression when the voltage of the I-component data signal is Vmi (where mi=0, 1, . . . , ni/2−1 when the number of the I-component data signal level is ni) and the half wavelength voltage of the I-component optical modulation unit is Vπi, anda progression {sin(Vmq/Vπq×π/2)} becomes an arithmetic progression when the voltage of the Q-component data signal is Vmq (where mq=0, 1, . . . , nq/2−1 when the number of the Q-component data signal level is nq) and the half wavelength voltage of the Q-component optical modulation unit is Vπq.

10. An optical modulation device that generates a QAM signal whose order is set to be constant, the device comprising: an I-component optical modulation unit that modulates a continuous optical signal for an 1-component of the QAM signal;a Q-component optical modulation unit that modulates a continuous optical signal for a Q-component of the QAM signal;a phase shift unit that shifts a phase of an optical signal on an input side and/or an output side for the I-component optical modulation unit and/or the Q-component optical modulation unit to adjust a phase difference between an output of the I-component optical modulation unit and an output of the Q-component optical modulation unit after the phase shift to π/2;a QAM signal generation unit that multiplexes the output of the I-component optical modulation unit and the output of the Q-component optical modulation unit after the phase shift to generate the QAM signal;an I-component data signal output unit that outputs an I-component data signal whose amplitude is controlled to be constant to the I-component optical modulation unit to cause the I-component optical modulation unit to modulate a continuous optical signal using the I-component data signal;a Q-component data signal output unit that outputs a Q-component data signal whose amplitude is controlled to be constant to the Q-component optical modulation unit to cause the Q-component optical modulation unit to modulate a continuous optical signal using the Q-component data signal;an I-component bias voltage signal output unit that outputs a signal having an I-component bias voltage corresponding to a null point of the I-component optical modulation unit to the I-component optical modulation unit;a Q-component bias voltage signal output unit that outputs a signal having a Q-component bias voltage corresponding to a null point of the Q-component optical modulation unit to the Q-component optical modulation unit;an I-component bias voltage signal adjustment unit that maximizes or minimizes an intensity of the QAM signal to adjust the I-component bias voltage to an optimum value; anda Q-component bias voltage signal adjustment unit that maximizes or minimizes the intensity of the QAM signal to adjust the Q-component bias voltage to an optimum value.

11. The optical modulation device according to claim 10, wherein Vmi is a constant at all times when a voltage of the I-component data signal is Vmi (where mi=0, 1, . . . , ni/2−1 when a number of the I-component data signal level is ni),Vmq is a constant at all times when a voltage of the Q-component data signal is Vmq (where mq=0, 1, . . . , nq/2−1 when a number of the Q-component data signal level is nq),the I-component bias voltage signal adjustment unit: adjusts the I-component bias voltage to maximize the intensity of the QAM signal at all times when a constant set by Equation 27 is negative, andadjusts the I-component bias voltage to minimize the intensity of the QAM signal at all times when the constant set by Equation 27 is positive,

12. A bias voltage control method, comprising: I and Q component optical modulation steps of modulating continuous optical signals for I and Q components of a QAM signal;a phase shift step of shifting a phase of an optical signal on an input side and/or an output side for the I-component optical modulation step and/or the Q-component optical modulation step and adjusting a phase difference between an output of the I-component optical modulation step and an output of the Q-component optical modulation step after the phase shift to π/2; anda QAM signal generation step of multiplexing the output of the I-component optical modulation step and the output of the Q-component optical modulation step after the phase shift to generate the QAM signal, whereinthe I-component optical modulation step sequentially includes an I-component signal input step of inputting an I-component data signal used to modulate a continuous modulation signal, andinputting a signal having an I-component bias voltage corresponding to a null point of the I-component optical modulation step, andan I-component bias voltage signal adjustment step of: determining whether an intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the I-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the I-component bias voltage,adjusting the I-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, andadjusting the I-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time, and whereinthe Q-component optical modulation step sequentially includes:a Q-component signal input step of inputting a Q-component data signal used to modulate a continuous modulation signal, andinputting a signal having a Q-component bias voltage corresponding to a null point of the Q-component optical modulation step, anda Q-component bias voltage signal adjustment step of:determining whether the intensity of the QAM signal at a drift non-occurrence time where no drift occurs in the Q-component bias voltage becomes larger or smaller than the intensity of the QAM signal at a drift occurrence time where a drift occurs in the Q-component bias voltage,adjusting the Q-component bias voltage to maximize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes larger than the intensity of the QAM signal at the drift occurrence time, andadjusting the Q-component bias voltage to minimize the intensity of the QAM signal when determining that the intensity of the QAM signal at the drift non-occurrence time becomes smaller than the intensity of the QAM signal at the drift occurrence time.