JP5099506B2 - Optical QAM signal demodulation method and demodulator - Google Patents

Description

  The present invention relates to an optical QAM signal demodulation method, a demodulation device, and the like. More specifically, the present invention relates to a method for demodulating an optical QAM signal, a demodulating device, and the like that can dynamically change a demodulation boundary in accordance with a peak position of a received signal.

  DQPSK (differential quadrature phase-shift-keying), QPSK (quad phase shift-keying: quadrature phase-shift-keying), APSK (amplitude-and-phase-Q) (amplitude-and-phase-modulation) Various multilevel optical modulation formats such as quadrature-amplitude-modulation have been studied for the purpose of improving the frequency utilization efficiency of optical communication channels.

  In particular, DQPSK using two parallel Mach-Zehnder modulators (DPMZM) is the most realistic and attracts attention. This is because the in-phase component (I component, in-phase components) and the quadrature phase component (Q component, quadrature components) of the carrier signal can be separately modulated by using the modulator. A DQPSK signal generator capable of generating an optical modulation signal exceeding 100 Gb / s by using two high-speed binary data signals having a modulation speed of 50 Gbaud or more has been developed.

  QAM (Quadrature Amplitude Modulation) has higher spectral efficiency than QPSK. The performance of the QAM scheme can be enhanced by using DQPSK using the above-described DPMZM.

  In ideal QAM (Quadrature Amplitude Modulation), the in-phase and quadrature components of the optical amplitude are neatly arranged, so it can be demodulated only by A / D conversion after separating the in-phase and quadrature components. . However, in practice, the optical QAM signal is distorted. This is caused by misalignment of elements constituting the modulator / demodulator, variation in characteristics, nonlinear optical effect of the transmission path, and the like. It is difficult to control these elements that distort the optical QAM signal. Therefore, a method capable of demodulating the optical QAM signal is desired even when the optical QAM signal is distorted.

  Japanese Patent Laying-Open No. 2003-101602 (Patent Document 1 below) discloses a wireless QAM signal decoding apparatus. This wireless QAM signal decoding apparatus determines a multilevel QAM signal by changing the threshold values of the I component and the Q component. More specifically, the wireless QAM signal decoding apparatus compares the I component of the multilevel QAM signal with the I axis threshold value to determine the I axis area, and divides the I component by the determination result to determine the I axis A division value is obtained, and the I-axis threshold is updated based on the division value for I-axis. The Q component of the multi-level QAM signal is compared with the Q-axis threshold value to determine the Q-axis area, and the Q component is divided by the determination result to obtain the Q-axis divided value. The Q-axis threshold is updated based on the above. The wireless QAM signal decoding apparatus decodes the wireless QAM signal using the threshold value thus obtained.

  On the other hand, the factor that distorts the optical QAM signal changes with time. However, the wireless QAM signal decoding device disclosed in Japanese Patent Application Laid-Open No. 2003-101602 has a problem in that it cannot always follow a change in time because the threshold value is fixed. In addition, there is a problem that it is not always sufficient to correct the distortion of the optical QAM signal.

  Japanese Patent Application Laid-Open No. 2007-110386 (Patent Document 2 below) calculates an average value of absolute values of each data of symbol points, and based on the average value, a threshold value between symbols of the I signal and the Q signal. An apparatus for decoding a wireless QAM signal for obtaining the above is disclosed. However, the wireless QAM signal decoding device disclosed in the publication has a problem that it is not always sufficient to correct distortion of the optical QAM signal.

In particular, a wireless QAM signal and an optical QAM signal have completely different factors for distorting the QAM signal, and thus the QAM signal is distorted differently. Therefore, there is a problem that the distortion of the optical QAM signal cannot always be corrected appropriately even if the conventional radio QAM signal demodulator is used as it is.
JP 2003-101602 A Japanese Patent Laid-Open No. 2007-110386

  An object of this invention is to provide the demodulation method and apparatus which can correct | amend distortion of a signal appropriately.

  An object of the present invention is to provide a method and apparatus for demodulating an optical QAM signal that can appropriately correct distortion of the optical QAM signal.

  In the conventional radio QAM signal demodulation method, although the threshold values of the I component and the Q component are changed, the boundary is set in a lattice shape. For this reason, it is considered that the method cannot sufficiently cope with demodulation of optical multilevel signals such as optical QAM signals. Therefore, the present invention basically obtains a histogram of the optical multilevel signal on the constellation map, extracts the peak position, and flexibly changes the shape of the boundary based on the peak position. This is based on the knowledge that an optical multilevel signal including distortion can be effectively demodulated. Specifically, the present invention relates to the following inventions.

  According to a first aspect of the present invention, a peak position calculating step for obtaining a peak position of a histogram of a received signal in a constellation map, and a signal demodulation based on the peak position obtained in the peak position calculating step. The present invention relates to a signal demodulating method including a boundary calculating step for obtaining a boundary of the signal and a decoding step for decoding a received signal using the boundary obtained in the boundary calculating step.

  In a preferred aspect of the first aspect of the present invention, the received signal is a QAM signal. The received signal is an optical signal having a value of 2 or more, and the received signal is an optical QAM signal, an optical ASK signal, an optical PSK signal, an optical FSK signal, an optical CPFSK signal, an optical APSK signal, or Any of optical MSK signals may be used. The most preferred signal is an optical QAM signal.

  In a preferred aspect of the first aspect of the present invention, the boundary calculation step is a peak obtained in the peak position calculation step, and a diagonal point for obtaining a diagonal point of a quadrangle formed by four adjacent ones. A calculation step, a peak obtained in the peak position calculation step, and a peak located in the outer contour, a midpoint calculation step for obtaining a midpoint between adjacent peaks, and a plurality of points obtained in the diagonal point calculation step And a point group connecting step of connecting a plurality of midpoints obtained in the midpoint calculating step and adjacent ones with the connecting line as a boundary.

  In a preferred aspect of the first aspect of the present invention, the boundary calculation step is a peak obtained in the peak position calculation step, and a diagonal point for obtaining a diagonal point of a quadrangle formed by four adjacent ones. A plurality of diagonal points obtained in the calculation step and the diagonal point calculation step, which connect adjacent ones, and for those diagonal points located on the outer side, move the connecting line in the outer direction. And a point group connection / extension process with the connection line and the extension line as a boundary.

  In a preferred embodiment of the first aspect of the present invention, the boundary calculation step is a peak obtained in the peak position calculation step, and a midpoint calculation step for obtaining a midpoint of adjacent ones, and the midpoint calculation step Connect the adjacent ones of the midpoints obtained in step 1 above, and for those located outside the midpoint, extend the connecting line in the outer direction and use the connecting line and the extended line as the boundary. Point cloud connection / extension process.

  In a preferred embodiment of the first aspect of the present invention, the peak position calculating step includes a preprocessing step. The preprocessing step includes an amplitude correction step, a weighting step, a fourth power calculation step regarding the received signal, a fourth power value average calculation step, and a phase shift step. The amplitude correction step is a step of correcting the amplitude of the in-phase component corresponding to the peak position in the constellation map and correcting the amplitude of the quadrature component. The weighting step is a step of weighting the corrected in-phase component and quadrature component. The fourth power calculation step for the received signal is a step for obtaining a value related to the phase drift by performing a fourth power operation on the weighted peak value. The mean value calculation step of the fourth power value is a step of obtaining a time average value of values related to the phase drift. The phase shift step is a step of obtaining a phase shift value based on the time average value, shifting the phase of the received signal by a predetermined amount based on the phase shift value, and obtaining a peak after phase shift correction. The boundary calculating step is a step of determining the boundary based on the peak position after the phase shift correction. This can be used in appropriate combination with the demodulation method described above.

  A preferred embodiment of the first aspect of the present invention relates to the demodulation method according to any one of the above, wherein the weighting step of the received signal is a step of weighting the received signal using an exponential function.

  A preferred embodiment of the first aspect of the present invention relates to the demodulation method according to any one of the above, wherein the received signal weighting step is a step of extracting received signals located at four corners of a constellation map.

  In a preferred aspect of the first aspect of the present invention, the weighting step of the received signal includes a step of extracting a received signal located at the four corners of the constellation map and a step of extracting the received signal located at the center of the constellation map. And the step of weighting so that the peaks located at the four corners of the extracted constellation map become large and the step of weighting so that the peak located at the center of the extracted constellation map becomes large The present invention relates to any one of the demodulation methods.

  In a preferred aspect of the first aspect of the present invention, the boundary calculation step further includes a step of providing a dummy symbol in a constellation map, and the boundary calculation step includes a position of the dummy symbol and a peak position calculation step. The present invention relates to the demodulation method according to any one of the above, wherein the boundary is obtained based on the obtained peak position.

  In a preferred embodiment of the first aspect of the present invention, the step of providing a dummy symbol in the constellation map includes a step of obtaining an inner product of two peaks in the constellation map, and an inner product value obtained in the step of obtaining the inner product is predetermined. A step of determining whether the inner product value is within a predetermined range, a step of obtaining a sum vector of two peaks obtained from the inner product, and installing a dummy peak at a position corresponding to the sum vector The demodulation method according to any one of the above, wherein

  In a preferred aspect of the first aspect of the present invention, the boundary calculation step is a peak obtained in the peak position calculation step, and a diagonal point for obtaining a diagonal point of a quadrangle formed by four adjacent ones. Including a calculation step and a point group connection step that connects a plurality of diagonal points obtained in the diagonal point calculation step and has a connecting line as a boundary, and the diagonal point calculation step is the same in the constellation map. The demodulation method according to any one of the above, wherein only the diagonal points belonging to the quadrant are obtained.

  In a preferred aspect of the first aspect of the present invention, the boundary calculation step includes a diagonal point calculation step, a midpoint calculation step, and a point group connection step, and the diagonal point calculation step includes the peak position calculation step. Is a step of obtaining a diagonal point of a quadrangle formed by four adjacent objects, and the midpoint calculation step in the previous period is a peak obtained in the peak position calculation step, Is a step of obtaining a midpoint between adjacent peaks for the peak located at a point, and the point cloud connection step is obtained by a plurality of diagonal points obtained in the diagonal point calculation step and the midpoint calculation step. By connecting adjacent diagonal points and midpoints that are multiple midpoints, and connecting diagonal points within a certain distance from the midpoint using a curve, a connection point is obtained, The recovery according to any one of the above, which is a step using the obtained connecting line as a boundary. A method for.

  According to a second aspect of the present invention, there is provided a peak position calculating means for obtaining a peak position of a histogram of a received signal in a constellation map, and a signal demodulating based on the peak position obtained by the peak position calculating means. Boundary calculating means for obtaining a boundary; and decoding means for decoding a received signal using the boundary obtained by the boundary calculating means, wherein the boundary calculating means includes a peak obtained by the peak position calculating means. And a diagonal point calculating means for obtaining a diagonal point of a quadrangle formed by four adjacent objects, and a peak obtained by the peak position calculating means between adjacent peaks. A midpoint calculation means for obtaining a midpoint, a plurality of diagonal points obtained by the diagonal point calculation means, and a plurality of midpoints obtained by the midpoint calculation means, which are adjacent to each other Includes a point group connecting means bounded by connecting lines, and for the signal of the demodulator.

  In a preferred embodiment of the second aspect of the present invention, the received signal is an optical signal having a binary value or more, and the received signal is an optical QAM signal, an optical ASK signal, an optical PSK signal, an optical signal. One of the FSK signal, the optical CPFSK signal, the optical APSK signal, and the optical MSK signal, and the optical QAM signal is most preferable.

  According to a third aspect of the present invention, there is provided a peak position calculation means for obtaining a peak position in a received signal constellation map, and a boundary for demodulating the signal based on the peak position obtained by the peak position calculation means. And a decoding means for decoding the received signal using the boundary obtained by the boundary calculating means, the boundary calculating means at the peak obtained by the peak position calculating means. A diagonal point calculating means for obtaining a diagonal point of a quadrangle formed by four adjacent objects, and a plurality of diagonal points obtained by the diagonal point calculating means for connecting adjacent ones. , With respect to the diagonal points located outside, including a point group connection / extension means extending the connection line in the outer direction and having the connection line and the extension line as a boundary. .

  In a preferred aspect of the second aspect of the present invention, the received signal is an optical QAM signal.

  ADVANTAGE OF THE INVENTION According to this invention, the demodulation method and apparatus which can correct | amend distortion of a signal appropriately can be provided.

  ADVANTAGE OF THE INVENTION According to this invention, the demodulation method and apparatus of an optical QAM signal which can correct | amend distortion of an optical QAM signal appropriately can be provided.

  FIG. 1 is a flowchart for explaining a signal demodulation method according to the present invention. As shown in FIG. 1, the signal demodulation method of the present invention basically includes a peak position calculation step (step 100), a boundary calculation step (step 200), a decoding step (step 300), including. For the sake of simplicity, the present invention will be described with a focus on the case of decoding an optical QAM signal. However, the scope of the present invention is not limited to an optical QAM signal processing method or processing apparatus.

Peak position calculation step (step 100)
The peak position calculation step is a step for obtaining the peak position in the received signal constellation map. The constellation map is also called a signal point arrangement diagram (for example, see Japanese Patent Application Laid-Open Nos. 2003-101602 and 2007-110386).

  The constellation map may be obtained by using a known photodetector for measuring an optical signal. Examples of the light detection device include a heterodyne detector capable of performing heterodyne detection, an A / D converter, and a control device for analyzing A / D converted digital information. The control device may be a computer. Such a computer includes a computer having an input / output unit, a control unit, a calculation unit, and a storage unit. These parts are connected by a bus or the like so that information can be exchanged. And the computer has what memorize | stored the program for functioning a computer as a demodulation apparatus of this invention in the memory | storage part. In addition, the arithmetic unit may include an electronic circuit that can perform arithmetic processing for causing a computer to function as the demodulator of the present invention. Further, an electronic circuit for digital information processing may be used as the control device. This electronic circuit can perform arithmetic processing for causing the control device to function as the demodulation device of the present invention. The peak position calculation step is realized by, for example, a peak position calculation means for obtaining a peak position of a received signal histogram in the constellation map.

  FIG. 2 is a graph instead of a drawing showing an example of the optical QAM signal on the constellation map. As shown in FIG. 2, the optical QAM signal passing through the transmission path is greatly distorted. That is, when the optical QAM signal is demodulated using a lattice-like boundary, it is predicted that it cannot be demodulated properly. Therefore, in the present invention, the peak position of the histogram in the constellation map is obtained by the peak position calculation means. In the following example, the peak position calculating means includes a histogram calculating means for obtaining a histogram that is the distribution of the optical QAM signal on the constellation map, and a peak position extracting means for extracting the peak position from the histogram obtained by the histogram calculating means. And have.

  FIG. 3 is a graph instead of a drawing showing an example of a histogram of the optical QAM signal based on the optical QAM signal on the constellation map of FIG. That is, in the present invention, for example, as shown in FIG. 3, a histogram of the optical QAM signal based on the optical QAM signal on the constellation map is obtained. Such a histogram can be easily obtained using a photodetector and the above-described control device, computer, or the like. Then, each vertex of the obtained histogram is extracted as a peak position. Methods for obtaining vertices from a histogram are known. The peak position can be easily extracted using, for example, a differentiation circuit.

  FIG. 4 is a diagram showing peak positions and diagonal points based on the optical QAM signal on the constellation map of FIG. As shown in FIG. 4, a peak is obtained based on the optical QAM signal on the constellation map of FIG. The peak positions are indicated by (A) to (P) in FIG. In FIG. 4, diagonal points to be described later are shown as (1a) to (1i).

  An example of the peak position calculation step includes a pre-processing step (step 110). The preprocessing step includes, for example, an amplitude correction step (step 111) for each of the in-phase component and the quadrature component of the received signal used for obtaining a peak, a weighting step (step 112) for each of the in-phase component and the quadrature component of the received signal, Examples include a fourth power calculation process (step 113) related to the received signal, a mean value calculation process (step 114), and a phase shift process (step 115). When phase drift does not occur in a QPSK signal or the like, the sum of the Q component multiplied by the imaginary unit and the I component is raised to one point for any state on the constellation map. This pretreatment process uses this property. That is, the time transition of the value obtained by raising each complex amplitude to the fourth power is used for phase correction. The amplitude correction process (step 111) of the received signal and the average calculation process of the fourth power value (step 114) are optional processes and can be omitted. After the pre-processing step, the boundary can be calculated through the boundary calculation step described above. The pretreatment process can be realized by a pretreatment device, for example. The preprocessing device includes, for example, a received signal amplitude correcting device, a received signal weighting device, a fourth power calculation device for a received signal, a fourth power value average calculating device, and a phase shift device.

Received signal amplitude correction step (step 111)
The received signal amplitude correcting step is a step for correcting the amplitudes of the I component and Q component of the received signal. That is, the sensitivity of the I component and the Q component varies depending on the receiver. Therefore, an appropriate constellation map can be obtained by correcting the I component and the Q component in the receiving apparatus. This process can be implemented using a received signal amplitude correction device. The received signal amplitude correction device, for example, grasps the characteristics of the reception device in advance and stores correction values for the I component and the Q component. Then, the received signal amplitude correction device multiplies each of the I and Q components of the signal received by the receiving device by this correction value to obtain a value after amplitude correction. In this way, variations in the I component and the Q component due to the receiving apparatus can be corrected.

Received signal weighting step (step 112)
The received signal weighting step is a step of weighting the received signal. The ratio between the optical signal and the quantum noise (S / N ratio) increases in proportion to the light intensity. Thus, an arithmetic process that increases the weight of an optical signal with a large absolute value is one of the weights. This arithmetic processing can be implemented by a peak weighting device. The received signal weighting device includes, for example, an exponential function table. The received signal weighting device reads an exponent value corresponding to the input peak value from the exponent function table. In this way, an exponent value corresponding to the absolute value of the amplitude can be obtained. In this example, the peak weighting device that performs the arithmetic processing using the exponential function table has been described. However, this calculation may use an exponent operation circuit or may be implemented by software. The weighting function is not limited to the exponential function. The I component and Q component that have been subjected to the weighting process are denoted as I _w and Q _w , respectively.

Examples of weighting using an exponential function are as follows.
I _w + jQ _w = (I + jQ) × exp (A)
A = (I ² + Q ² ) ^1/2

Fourth power calculation process for received signal (step 113)
The fourth power calculation process related to the received signal is a process for obtaining the phase drift by performing the fourth power calculation related to the received signal. Depending on the phase drift of the received signal, the declination of the fourth power of the complex number with the I component and Q component in the constellation map as the real part and imaginary part respectively changes. In the present invention, as described above, the weighted received signal is subjected to the fourth power calculation process to obtain the phase drift. In this way, the phase drift can be obtained for the optical signal.

Specifically, in the fourth power calculation step related to the received signal, the fourth power (z) related to the received signal is obtained according to the following equation.
z = (I _w + jQ _w ) ⁴
Arg (z / 4) corresponds to the phase drift.

The fourth power calculation process related to the received signal can be implemented using a fourth power calculation device related to the received signal. This fourth power calculator can be easily implemented using, for example, an adder and a fourth power calculation table. Then, _{I w} and jQ _w is inputted to the fourth power calculation device. The fourth power calculator calculates I _w + jQ _w using an adder. Next, the fourth power calculation device obtains (I _w + jQ _w ) ⁴ by referring to the fourth power calculation table. In this way, the z value can be obtained. Of course, this fourth power calculator may be mounted using a circuit different from the above. Further, this fourth power calculator may be implemented by software.

Average calculation process of the fourth power value (step 114)
The average calculation process of the fourth power value is a process for obtaining a time average value of z values. The time average value of this z value is expressed as z _av . By obtaining the time average value of z values, time fluctuation can be suppressed. In this operation, the z value is obtained several times and the average is obtained. That is, the above steps 111 to 113 are repeated n times to obtain a plurality of z values. Then, add the obtained z values. Then, the added z value may be divided by n. n is generally 2 or more and 1000 or less. The average calculation process of the fourth power value can be implemented using a fourth power value average calculation device. This average calculating device can be easily implemented using a storage device, an adder circuit and a multiplier circuit. The storage device stores a plurality of z values and 1 / n values. The adder circuit adds a plurality of z values. Thereby, this average calculation apparatus obtains Σz. The multiplication circuit multiplies Σz by 1 / n. Thereby, this average calculation apparatus obtains Σz / n. This Σz / n is _{z av.} Of course, the mean value calculation device for the fourth power value may be mounted using a circuit different from the above. Further, this mean value calculation device for the fourth power value may be implemented by software.

Phase shift process (step 115)
In the phase shift step, the phase of the received signal is corrected by shifting the phase by a predetermined amount, and the corrected received signal is obtained. In the phase shift process, the amount of phase to be shifted can be obtained using the previously obtained value. Specifically, Arg (z _av / 4) is given as the amount of phase shift in the phase shift process. The phase shift device can be obtained using the Arg table. This Arg table is an improved Arg table in which z _av and a corresponding Arg (z _av / 4) are stored in association with each other. That _is, when _{z av} is input, the phase shift device _reads the Arg _(z av / 4) from Arg table using the _{z av.} In this way, the phase shift device obtains the phase shift amount. Then, the peak position in the constellation map of the received signal is shifted.

  The above preprocessing is particularly effective when demodulating a QAM signal. FIG. 7A is a diagram illustrating an example of a constellation map of a QAM signal. FIG. 7B is a diagram showing a constellation map when preprocessing without weighting is performed on the signal of FIG. 7A. FIG. 7C is a diagram showing a constellation map when preprocessing for weighting with an exponential function is performed on the signal of FIG. 7A. From FIG. 7B, it is possible to process the QAM signal in a state where it can be decoded according to the present invention by performing preprocessing without weighting. Also, it can be seen from FIG. 7C that the symbols of the received signal are more clearly separated by performing preprocessing with weighting. Therefore, it can be seen that preprocessing using weighting can be performed more appropriately.

  Of the boundary calculation step (step 100), another embodiment (pattern) different from the above includes a pre-processing step (step 120). The pre-processing step (step 120) includes, for example, a received signal amplitude correcting step (step 121), a received signal weighting step (step 122), a fourth power calculation step (step 123) relating to the received signal, and a fourth power value. An average calculation step (step 124) and a phase shift step (step 125). The amplitude correction step (step 121) and the fourth power value average calculation step (step 124) are optional steps and can be omitted. After the pre-processing step, the boundary can be calculated through the boundary calculation step described above. The pretreatment process can be realized by a pretreatment device, for example. The preprocessing device includes, for example, a received signal amplitude correcting device, a received signal weighting device, a fourth power calculation device for a received signal, a fourth power value average calculating device, and a phase shift device.

  The received signal amplitude correction step (step 121), the fourth power calculation step for the received signal (step 123), the fourth power average calculation step (step 124), and the phase shift step (step 125) are: The same processing as described above is performed. Therefore, the description of these steps is cited and the description is omitted.

Received signal weighting step (step 122)
The received signal weighting step is a step of weighting the received signal. Unlike the case of a QPSK signal, in the case of a QAM signal, if the fourth power calculation process related to the received signal is applied as it is, even if there is no phase drift, it is obtained by the fourth power calculation process depending on the position of the symbol on the constellation map. The point is different. Specifically, the fourth power is calculated using symbols that exist on a straight line that passes through the origin of the constellation map and has an inclination of ± 45 degrees with respect to the horizontal axis (I axis) and symbols that do not exist on the straight line. The point obtained by a process differs. In order to suppress this influence, it is necessary to perform an average calculation process of the fourth power value, which is a subsequent process, on a sufficiently large number of received signals, which increases the amount of calculation. In order to avoid this, calculation processing is performed such that a symbol used in the fourth power calculation process is selected by weighting the received signal. This arithmetic processing can be implemented by a weighting device for received signals.

  One of the weighting methods for the 16QAM signal is weighting for selecting a symbol that forms a quadrangle with the largest area and a quadrangle with the smallest area among symbols on the constellation map. Specifically, referring to FIG. 4, point (A), point (D), point (M) and point (P), point (F), point (G), point (J) and point in FIG. (K). Weighting is performed so as to increase the weights of these eight points. From the viewpoint of S / N, it is desirable to calculate to the fourth power only at the four corners without including the central four peaks. However, only 1/4 of all data can be used for phase drift correction. In that case, when the modulation rate is slow, there is a risk that phase drift correction in a time zone where the four corner points do not come out will not be successful. Therefore, the vectors at the fourth power are in the same direction, and the central four peaks are also included in the fourth power calculation. As an example of the weighting function, there is a combination of rectangular functions that return 0 or 1 according to the distance from the center of gravity of 16 symbols. As a result, the phase drift value can be obtained with high accuracy.

Boundary calculation process (step 200)
The boundary calculation step is a step for obtaining a boundary for demodulating the signal based on the peak position obtained in the peak position calculation step. This step is realized by the boundary calculation means for obtaining the boundary for demodulating the signal based on the peak position obtained by the peak position calculation means.

  As an embodiment (pattern) having a boundary calculation step (step 200), a diagonal point calculation step (step 201), a midpoint calculation step (step 202), and a point group connection step (step 203) are included. Can be given. Of course, the order of the diagonal point calculation step (step 201) and the midpoint calculation step (step 202) may be reversed.

  This process can be realized by, for example, a boundary calculation device including a diagonal point calculation device, a midpoint calculation device, and a point group connection device. The diagonal point calculation device determines the diagonal points of the quadrangle formed by the four adjacent peaks, which are the peaks obtained by the peak position calculation means. The work for obtaining the diagonal point by the diagonal point calculation device may be performed by reading the coordinates of the four peaks and performing a known calculation process. The midpoint calculation device obtains a midpoint between adjacent peaks for a peak located outside the peak obtained by the peak position calculation means. The midpoint calculation device reads the coordinates of the peaks located on the outer contour, and thereby determines adjacent peaks. Then, the midpoint of these is obtained from the coordinates of adjacent peaks. The point cloud connecting device connects a plurality of diagonal points and a plurality of midpoints adjacent to each other. Adjacent diagonal points and adjacent diagonal points and midpoints can be easily determined using their coordinate values. After judging adjacent points, the boundary can be obtained by connecting them.

Diagonal point calculation step (step 201)
The diagonal point calculating step (step 201) is a step of determining the square diagonal points formed by the four adjacent points, which are the peaks obtained in the peak position calculating step. Four adjacent peaks can be easily analyzed by a control device connected to the photodetector. For example, the vertex (A), the vertex (B), the vertex (E), and the vertex (F) are four adjacent peaks. The square diagonal points formed by these vertices can also be easily obtained by the control device. In this case, the diagonal point is (1a).

Midpoint calculation step (step 202)
The midpoint calculation step (step 202) is a step of obtaining a midpoint between adjacent peaks for the peaks obtained in the peak position calculation step and located on the outer contour.

  FIG. 5 is a diagram showing peak positions and midpoints based on the optical QAM signal on the constellation map of FIG. The peak positions are indicated by (A) to (P) in FIG. In FIG. 5, the midpoints are shown as (2a) to (2l). Peaks located on the outer shell and adjacent peaks can be easily analyzed by a control device connected to the photodetector. For example, the vertex (A) and the vertex (B) are peaks located in adjacent outlines. The midpoint of each of these vertices can also be easily determined by the control device. In this case, the midpoint of the vertex (A) and the vertex (B) is the midpoint (2a).

Point cloud connection step (step 203)
In the point group connection step (step 203), a plurality of diagonal points obtained in the diagonal point calculation step and a plurality of midpoints obtained in the midpoint calculation step are connected to each other, and a connecting line is obtained. This is the process of making a boundary. When the midpoint and the diagonal point are connected, the boundary is obtained by extending from the diagonal point toward the midpoint.

  FIG. 6 is a graph replaced with a drawing in which the boundary is obtained based on the optical QAM signal on the constellation map of FIG. That is, the boundary is obtained using the connection points connected in the point group connection process.

  Of the boundary calculation step (step 200), another embodiment (pattern) different from the above includes a diagonal point calculation step (step 211) and a point group connection / extension step (step 212). . This step can be realized by a boundary calculation device including a diagonal point calculation device and a point group connection / extension device. The diagonal point calculation device obtains a square diagonal point formed by four adjacent peaks. The point cloud connection / extension device connects adjacent diagonal points. Further, the point cloud connecting / extending device extends the connecting line in the outline direction for diagonal points located outside the outline.

  The diagonal point calculating step (step 211) is the same as the diagonal point calculating step (step 201) described above. In this embodiment, a plurality of diagonal points obtained in the diagonal point calculating step, which connect adjacent ones, and extend diagonal lines of the diagonal points located in the outer direction in the outer direction. However, the connecting line and the extension line are the boundaries.

  That is, for example, the diagonal point (1a) and the diagonal point (1b) in FIG. 4 are connected, and the diagonal point (1b) and the diagonal point (1c) are connected. In this way, a plurality of diagonal points adjacent to each other are connected. On the other hand, the diagonal point (1a) is located outside the diagonal point. Therefore, the connecting line between the diagonal point (1a) and the diagonal point (1b) is extended in the direction from the diagonal point (1b) to the diagonal point (1a). In the same manner, the other diagonal points located on the outline extend the connecting line in the outline direction. In this embodiment, the connection line and the extension line thus obtained are used as boundaries. That is, in this embodiment, since it is not necessary to perform a calculation process for obtaining a midpoint, the processing speed can be increased.

  In the point group connection / extension process (step 212), a plurality of diagonal points obtained in the diagonal point calculation process, which connect adjacent ones and which are located outside the diagonal points, In this process, the connecting line is extended in the outer direction, and the connecting line and the extended line are used as a boundary. Specifically, the process is as follows. First, a plurality of diagonal points obtained in the diagonal point calculating step are read out. Then, the adjacent ones are selected from the read coordinate values of the diagonal points, while the ones located in the outline are determined from the read coordinate values of the diagonal points. Then, adjacent diagonal lines are connected. In addition, when connecting a diagonal point located on the outline and other diagonal points, the line connected to the diagonal point located on the outline is extended.

  Of the boundary calculation step (step 200), an embodiment (pattern) different from the above includes a midpoint calculation step (221) and a point group connection / extension step (222). This boundary calculation step can be achieved by the midpoint calculation means and the point group connection / extension means.

Midpoint calculation step (221)
The midpoint calculating step (221) is a step for determining the midpoint of the peak obtained in the peak position calculating step and adjacent. Although not specifically shown, for example, in FIG. 4, peak (A) and peak (B), peak (E) and peak (F), peak (I) and peak (J), peak (M) and peak (N) Are adjacent peaks. In this process, the midpoint of them may be obtained. The method for obtaining the midpoint can be easily obtained by obtaining the average of the coordinates of the peak positions.

Point cloud connection / extension process (222)
In the point group connection / extension step (222), a plurality of midpoints obtained in the midpoint calculation step are connected, and adjacent points are connected. This is a process of extending in the outer direction and using the connecting line and the extended line as a boundary. In other words, the midpoint obtained in the previous step may be connected and extended.

  Of the boundary calculation step (step 200), another embodiment (pattern) different from the above includes a step of providing a dummy symbol in the constellation map (step 251). After the step of providing the dummy symbol, the boundary can be calculated through the boundary calculation step described above. The step of providing the dummy symbol can be realized by a dummy symbol installation device, for example. The process of providing the dummy symbol is particularly effective when the signal to be decoded has a peak only on the I axis and the Q axis, such as an APSK signal.

  FIG. 8 is a conceptual diagram showing a constellation map of the APSK signal. As shown in FIG. 8, the APSK signal has peaks on the I axis and the Q axis. Therefore, although the decoding method of the present invention can be used, it cannot always be accurately decoded.

  FIG. 9 is a conceptual diagram showing a constellation map after dummy symbols are installed. As shown in FIG. 9, a constellation map such as a QAM signal can be obtained by installing dummy symbols. For this reason, it can decode effectively using the decoding method of this invention.

  That is, the step of providing the dummy symbol is a step of providing peaks corresponding to the peaks in the I axis and the Q axis in portions other than the I axis and the Q axis. The position where the dummy symbol is provided may be determined in advance. Ideally, the positions of the dummy symbols may be obtained using the positions of the peaks located on the I axis and the Q axis. That is, information on lattice points such as lattice point intervals may be obtained from peak position information, and dummy symbols may be placed at corresponding positions.

  Further, the process of providing the dummy symbol can be realized as follows, for example. That is, the inner product of the position vectors of the (n−1) th peak and the nth peak in the constellation map is calculated. If the inner product is not 0, both peaks are on the same axis. In this case, no dummy symbol is generated. On the other hand, when the inner product is 0, the peak is on a different axis. In this case, the sum vector of both peaks is obtained, and the position of this sum vector is set as the peak position of the dummy symbol. In this way, a dummy symbol can be obtained based on the observed peak.

  That is, the process of providing the dummy symbol can be achieved by a process including an inner product value calculating process and a sum vector calculating process. The step of providing a dummy symbol can be achieved by a dummy symbol installation device. Examples of the dummy symbol setting device include an inner product value calculating device and a sum vector calculating device.

  Of the boundary calculation step (step 200), another embodiment (pattern) different from the above includes a diagonal point calculation step (step 231) and a point group connection step (step 232). Note that a midpoint calculation step may be provided before the diagonal point calculation step or before the point group connection step.

  In this diagonal point calculation step (step 231), only diagonal points belonging to the same quadrant are obtained. That is, when there is an ideal peak only on the axis, such as an APSK signal, the decoding method described above cannot always be used efficiently. Therefore, in this aspect, the boundary is obtained using diagonal points only in the same quadrant.

  FIG. 10 is a diagram illustrating an example in which a boundary is obtained using only diagonal points belonging to the same quadrant. If this boundary is used, information on only the relative position (phase) relative to the origin can be obtained effectively. Information on the amplitude can be obtained using a known photodetector. For this reason, it is preferable that the decoding method of this aspect obtains phase information using this method and obtains amplitude information by a known method.

  As an embodiment (pattern) different from the above in the boundary calculation step (step 200), a diagonal point calculation step (step 241), a midpoint calculation step (step 242), and a point group connection step (step 243), , Which includes Of course, the order of the diagonal point calculation step (step 241) and the midpoint calculation step (step 242) may be reversed.

  In this aspect, when connecting the diagonal points in the point group connecting step (step 243), the diagonal points are connected using a circle.

  FIG. 11 is a diagram illustrating an example in which the boundary in the APSK signal is obtained. The boundary calculation process described above connects diagonal points. When the boundary of the APSK signal is calculated using such a method, the result is as shown in FIG. On the other hand, in the point group connection step (step 243), for example, four diagonal points having the same distance from the midpoint are obtained. Then, find the curves that pass through the diagonal and midpoints and connect them. This curve group is used as a boundary.

Decryption process (step 300)
The decoding step (step 300) is a step of decoding the received signal using the boundary obtained in the boundary calculation step. Since a value corresponding to the region surrounded by the boundary is assigned, decoding can be performed using information on which region the optical QAM signal is observed as. Such a decoding process can be realized using known software or an arithmetic circuit.

FIG. 12 is a diagram illustrating an example of processing for decoding a signal. In FIG. 12, in order to determine the presence / absence of a signal in a certain area, the area is divided into an area indicated by vector A ₁ and vector B ₁ and an area indicated by vector A ₂ and vector B ₂ . When the vector from the base point of the vector A ₁ and the vector B ₁ to the observation point of the signal is a vector p, if the following expression 1 is satisfied, the signal is a triangle having the vector A ₁ and the vector B ₁ as two sides. It exists in the area | region or boundary. Hereinafter a _1, if B ₁ is not satisfied the following condition, a similar determination was performed using a combination of vector A ₂ and the vector B _2, a triangular area to the vector A ₂ and the vector B ₂ two sides The presence / absence of a signal at is determined. This determination is repeated for all the areas defined by the boundary lines defined by the above-described procedure, and it is determined which area on the constellation map contains the signal, that is, decoding.

  The present invention can be preferably used as an information communication system including a signal generator and the signal demodulator described above. Hereinafter, examples of the signal generation device and the signal generation method in the present invention will be described.

  FIG. 13 is a schematic diagram of a quadrature amplitude modulation (QAM) signal generator used in the present invention. As shown in FIG. 13, the QAM signal generator of the present invention includes a first waveguide (2) and a first four-phase phase shift keying (QPSK) provided in the first waveguide (2). ) Provided in the signal generator (3), the second waveguide (5) having the first waveguide (2) and the multiplexing point (4), and the second waveguide (5) A second QPSK signal generator (6). This QAM signal generator is preferably one that branches from one input path and allows light to propagate to the first waveguide (2) and the second waveguide (5).

  A quadrant in which the QAM signal appears in a QPSK signal having a large amplitude (in FIG. 14, which region is sandwiched between the axes) is determined, and a position in the quadrant is determined by a small QPSK signal. In other words, the position of the black point in FIG. 14 is determined by the QPSK signal having a large amplitude, and the position of the white point (the position of the QAM signal) is determined by the QPSK signal having a small amplitude. Each QPSK signal can be obtained without using a multi-level electric signal.

The waveguide is a path through which an optical signal propagates, and examples thereof include a titanium diffusion waveguide (LiNbO ₃ waveguide) provided on a LiNbO ₃ substrate.

  As shown in FIG. 13, the first QPSK signal generator (3) and the second QPSK signal generator (6) include a main Mach-Zehnder waveguide (7a, 7b) and the main Mach-Zehnder waveguide, respectively. One having two sub Mach-Zehnder waveguides (8a, 9a, 8b, 9b) provided on both arms of the waveguide (7a, 7b) can be mentioned. The configuration of this QPSK signal generator is known as an optical SSB modulator, an optical FSK signal generator, or the like. Therefore, this QPSK signal generator can adopt a known method as appropriate. For an optical SSB (Single Side-Band) modulator, see, for example, Japanese Patent Laid-Open No. 2005-274806. For optical FSK signal generators with improved optical SSB modulators, see Japanese Patent Application Laid-Open Nos. 2007-86207, 2007-57785, and 2005-134897. These documents are incorporated herein by reference.

  The second QPSK signal generator (6) preferably outputs a signal having a smaller amplitude than the output signal from the first QPSK signal generator. The amplitude of the output signal from the second QPSK signal generator (6) is preferably adjusted to be half the amplitude of the output signal from the first QPSK signal generator. In this way, QAM signals with equal intervals can be obtained. Note that the half amplitude does not mean a half in the strict sense, but may be a value that can be substantially distinguished as a QAM signal. Specifically, the amplitude of the output signal from the second QPSK signal generator (6) is 30% to 70% of the amplitude of the output signal from the first QPSK signal generator. % Or more and 60% or less, or 45% or more and 55% or less. The above adjustment is achieved by, for example, an amplitude adjustment mechanism. Specifically, the amplitude adjusting mechanism is achieved by voltage control means for controlling the voltage applied to the bias electrode or the like. Further, the amplitude may be adjusted by appropriately providing an intensity modulator in the waveguide and performing predetermined intensity modulation by the intensity modulator.

  On the other hand, the phase difference of a QPSK signal or a 4-value ASK signal described later is 45 degrees (more precisely, 45 degrees + 90 n degrees (n is an integer), the same applies hereinafter), or 30 degrees, 22.5 degrees, 18 degrees. In cases other than an integral multiple of 90 degrees, such as 15 degrees, the second QPSK signal generator (6) outputs a signal having the same amplitude as the output signal from the first QPSK signal generator, It doesn't matter. In the following, a case where the phase difference of the QPSK signal is 0 degree, 90 degrees, 270 degrees or 180 degrees and a case where the phase difference of the quaternary ASK signal is 90 degrees or 270 degrees will be described. Alternatively, even if the phase difference of the quaternary ASK signal is other than these, it can be adjusted in the same manner.

  In order to make the amplitude of the output signal from the second QPSK signal generator (6) smaller than the amplitude of the output signal from the first QPSK signal generator, for example, the second QPSK signal generator A combination of an attenuator and an intensity modulator may be used as the detector, and driving may be performed to reduce the amplitude. Further, for example, the ratio of the intensity branching of the waveguide may be asymmetrical without using an attenuator. For example, when an optical signal is propagated from the input port to the first waveguide (2) and the second waveguide (5) through a branch point, the branch ratio is set to 4: 1. Good. Such adjustment of the branching ratio is well known and can be easily achieved by adjusting the coupler and the waveguide.

  A preferred embodiment of the signal generator is the main Mach-Zehnder waveguide (7a) of the first QPSK signal generator (3) and the main Mach-Zehnder waveguide (7b) of the second QPSK signal generator (6). , A first bias electrode (10a, 10b) and a second bias electrode (11a, 11b) for adjusting the bias voltage applied to each of the two arms constituting the main Mach-Zehnder waveguide, A third bias electrode (13a, 13b) is provided for applying a bias voltage to a signal propagating through both arms combined by the multiplexing unit (12a, 12b) of the main Mach-Zehnder waveguide (7a, 7b). And any one of the above-mentioned devices.

  By providing such a large number of bias electrodes, it is possible to effectively adjust the phase difference of signals propagating through the arms, thereby obtaining QAM signals in various modes.

In a preferred embodiment of the signal generator, the first QPSK signal generator (3) and the second QPSK signal generator (6) are provided on a LiNbO ₃ waveguide (21), and the first conductor is provided. A combining point (4) between the waveguide (2) and the second waveguide (5) is provided on the planar optical waveguide (PLC) (22) optically connected to the LiNbO ₃ waveguide (21). A device according to any one of the above. Thus, by performing hybrid integration of a substrate such as a lithium niobate substrate and a PLC, a signal generator including a plurality of MZMs (Mach-Zehnder modulators) can be manufactured relatively easily. The manufacturing method of the PLC is as disclosed in Japanese Patent Application Laid-Open No. 2006-195036.

  In other words, the QAM signal generator used in the present invention can be manufactured by combining the technology such as the optical SSB modulator and the optical FSK signal generator, the technology such as PLC, and the ordinary technical knowledge possessed by those skilled in the art. it can.

  If the optical QAM signal generator is used, an optical QAM signal can be obtained. A specific method for generating an optical QAM signal includes a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal generator (QPSK) signal generator provided in the first waveguide (2). 3) and a second waveguide (5) having a coupling point (4) between the first waveguide (2) and the second waveguide (5). A second QPSK signal generator (6) for outputting a signal having a smaller amplitude than the output signal from the QPSK signal generator, and the first QPSK signal generator (3) and the second QPSK signal generator. The signal generator (6) includes a main Mach-Zehnder waveguide (7a, 7b) and two sub Mach-Zehnder waveguides (7a, 7b) respectively provided on both arms of the main Mach-Zehnder waveguide (7a, 7b). 8a, 9a, 8b, 9b), quadrature amplitude modulation A QAM signal generation method using a QAM) signal generator, wherein the output signal is a half of the amplitude of the output signal from the first QPSK signal generator and the output signal from the first QPSK signal generator. A method for generating a QAM signal includes a step of combining the output signal from the second QPSK signal generator (6).

  As a preferred embodiment of the QAM signal generation method, the main Mach-Zehnder waveguide (7a) of the first QPSK signal generator (3) and the main Mach-Zehnder waveguide (7b) of the second QPSK signal generator (6) are used. ), The first bias electrode (10a, 10b) and the second bias electrode (11a, 11b) for adjusting the bias voltage applied to each of the two arms constituting the main Mach-Zehnder waveguide, respectively. And a third bias electrode (13a, 13b) for applying a bias voltage to a signal propagating through both arms to be combined at the multiplexing section (12a, 12b) of the main Mach-Zehnder waveguide (7a, 7b). ), And the four sub Mach-Zehnder waveguides (8a, 9a, 8b, 9b) The phase difference between the signals propagating through the two arms before the propagation signals are combined is 180 degrees, and each main Mach-Zehnder waveguide (7a, 7b) has a phase difference before the propagation signals are combined. , The phase difference between the signals propagating through the both arms is 90 degrees or 270 degrees, and the first QPSK signal generator (3) and the second QPSK signal generator (6) The phase difference between the signals propagating through the first waveguide (2) and the second waveguide (5) before being multiplexed at the multiplexing point (4) is 0 degrees, 90 degrees, and 270 degrees. Or a bias signal applied to the first bias electrode (10a, 10b), the second bias electrode (11a, 11b) and the third bias electrode (13a, 13b) so as to be 180 degrees. Control all or at least one , Method, and the like. The control of the phase difference is known, for example, as an optical SSB modulation technique or an optical FSK modulation technique, and can be achieved by appropriately using a known method. By adjusting in this way, a QAM signal can be obtained.

  As a preferred embodiment of the QAM signal generation method, a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal generator (3) provided in the first waveguide (2) And a second waveguide (5) having a coupling point (4) with the first waveguide (2), and the second waveguide (5), the first QPSK signal generation And a second QPSK signal generator (6) for outputting a signal having a smaller amplitude than the output signal from the detector, the first QPSK signal generator (3) and the second QPSK signal generator (6) is a main Mach-Zehnder waveguide (7a, 7b) and two sub-Mach-Zehnder waveguides (8a, 9a) respectively provided on both arms of the main Mach-Zehnder waveguide (7a, 7b). , 8b, 9b) A QAM signal generation method using a (QAM) signal generator, wherein a first ASK signal that is a four-value amplitude shift keying (ASK) signal is generated in the first main Mach-Zehnder waveguide (7a) In the second main Mach-Zehnder waveguide (7b), a second ASK signal which is a four-value amplitude shift keying (ASK) signal having a phase difference of 90 degrees or 270 degrees with the first ASK signal And a method of generating a QAM signal, including a step of combining the first ASK signal and the second ASK signal.

  As a preferable example of this aspect, the main Mach-Zehnder waveguide (7a) of the first QPSK signal generator (3) and the main Mach-Zehnder waveguide (7b) of the second QPSK signal generator (6) are: The first bias electrode (10a, 10b) and the second bias electrode (11a, 11b) for adjusting the bias voltage applied to each of the two arms constituting the main Mach-Zehnder waveguide, A third bias electrode (13a, 13b) for applying a bias voltage to a signal propagating through both arms combined in the multiplexing section (12a, 12b) of the Mach-Zehnder waveguide (7a, 7b); , The four sub-Mach-Zehnder waveguides (8a, 9a, 8b, 9b) have both arms before the propagation signals are combined. The phase difference of the transmitted signal is 180 degrees, and each main Mach-Zehnder waveguide (7a, 7b) has the phase difference of the signals propagating through the arms before the propagation signals are combined is 0 degree. , Or 180 degrees, and the first QPSK signal generator (3) and the second QPSK signal generator (6) are configured so that the propagation signal is multiplexed before being multiplexed at the multiplexing point (4). The first bias electrode (10a, 10b), the first bias electrode (10a, 10b), the phase difference between signals propagating through the first waveguide (2) and the second waveguide (5) are 90 degrees or 270 degrees. There is a method of controlling the bias signal applied to the second bias electrode (11a, 11b) and the third bias electrode (13a, 13b).

  That is, in this method, two quaternary ASK signals (with four symbols of +1, +1/3, −1/3, and −1) are generated, and these two are synthesized by the outermost interferometer. That's fine. At this time, the bias voltage of the outermost interferometer may be adjusted so that they have a phase difference of 90 degrees or 270 degrees.

  FIG. 15 is a conceptual diagram of a QAM signal generator using three QPSK signal generators arranged in parallel. As shown in FIG. 15, the QAM signal generator of this aspect includes a first waveguide (52) and a first four-phase phase shift keying (QPSK) provided in the first waveguide (52). ) Provided in a second waveguide (55) having a coupling point (54) between the signal generator (53) and the first waveguide (53), and the second waveguide (55). , A combining point (57) of the second QPSK signal generator (56) that outputs a signal having a smaller amplitude than the output signal from the first QPSK signal generator and the second waveguide (55). And a third waveguide (58) having a smaller amplitude than an output signal from the second QPSK signal generator, provided in the third waveguide (58). A quadrature amplitude modulation (QAM) signal generator comprising a QPSK signal generator (59) A.

  As the elements of the QAM signal generator of this aspect, those described above can be adopted as appropriate. In the QAM signal generator described above, it is basically possible to obtain a 16QAM signal by superimposing two 4ASK signals or superimposing two QPSK signals. On the other hand, the QAM generator shown in FIG. 15 can basically obtain a quadruple 64-QAM signal based on the same principle as described above. That is, a new QAM signal can be obtained by superimposing the center of a square having four symbols at the apexes on each white point in FIG.

That is, when n (n is a natural number equal to or greater than 2) QPSK signal generators are used, assuming that the amplitude of the nth QPSK signal has half the amplitude of the (n-1) th QPSK signal, 2 ²ⁿ QAM signals can be generated. This can be obtained by adding each of n QPSK signals with a phase difference of 0 degree, 90 degrees, 270 degrees, or 180 degrees. The amplitude of each QPSK may be adjusted so as to be halved sequentially, such as 1, 1/2, 1/4,... 1 / (2 ⁿ ).
Each level of the 2 ⁿ ASK signal is (2 ⁿ -1) / (2 ⁿ -1), (2 ^n-1 -2x1) / (2 ⁿ -1), ... (2 ^n-1 -2 * k ) / (2 ⁿ −1),... (2 ⁿ⁻¹ −2 * (2n−1)) / (2 ⁿ −1). Specifically, when obtaining a 64QAM signal, the levels of the ASK signal are 1, 5/7, 3/7, 1/7, -1/7, -3/7, -5/7, -1 respectively. Become. In order to adjust the intensity ratio of each QPSK, it can be adjusted in the same manner as described above. Specifically, an attenuator or the like may be installed in the optical circuit and the amplitude may be adjusted as appropriate. Further, the branching ratio may be adjusted when demultiplexing the input signal to each QPSK signal generator.

Automatic adjustment of bias It is necessary to maintain a phase difference of 90 degrees or 270 degrees when QPSK occurs. Since this is the same as the condition for generating an SSB modulation signal, a test sine wave signal is input to a bias electrode to which a phase difference of 90 degrees is desired at a low frequency (for example, several kilohertz or less) that does not affect transmission. When the SSB condition is satisfied, only one sideband component is output. However, when the SSB condition is shifted, both are output, and a double component of the modulation frequency is generated from these beats. If this is controlled to be minimized, a phase difference of SSB condition = 90 degrees can be obtained. The bias of the sub Mach-Zehnder waveguide can be set by inputting a sine wave signal and minimizing the same frequency component and maximizing the double component.

  Specifically, in FIG. 13, a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal generator (3) provided in the first waveguide (2). ) And the first waveguide (2), the second waveguide (5) having a coupling point (4), and the second waveguide (5), the first QPSK. A second QPSK signal generator (6) for outputting a signal having a smaller amplitude than the output signal from the signal generator, and the first QPSK signal generator (3) and the second QPSK signal. The generator (6) includes a main Mach-Zehnder waveguide (7a, 7b) and two sub-Mach-Zehnder waveguides (8a) respectively provided on both arms of the main Mach-Zehnder waveguide (7a, 7b). , 9a, 8b, 9b) and a test signal to each bias electrode A test signal applying device for applying the light, a light detecting device for detecting an output signal from the waveguide to which the test signal is applied by the test signal applying device, and detection data from the light detecting device, When the output spectrum detected by the detection device includes two sideband components, it has an automatic bias adjustment device that adjusts the bias voltage so as to reduce the double component of the modulation frequency. An amplitude modulation (QAM) signal generator can be mentioned. By doing so, it is possible to automatically adjust to the optimum bias situation.

  In the QAM signal generating apparatus according to this aspect, it is preferable that the test signal has a frequency of 10 kHz or less. That is, if the signal has a frequency of 10 kHz, the information transmission is not affected, so that the bias can be automatically adjusted without interfering with the information communication.

  Specifically, a predetermined signal is supplied to the main Mach-Zehnder waveguide, and the first sub-Mach-Zehnder waveguide is configured to reduce the amplitude of the same frequency component in the output light of the main Mach-Zehnder waveguide. And a bias adjustment method for adjusting a supply bias signal to the second sub Mach-Zehnder waveguide, wherein the predetermined signal supplied to the main Mach-Zehnder waveguide changes the phase of the optical signal by 180 degrees. There is a bias adjustment method that is an AC signal having an amplitude equal to or larger than the half-wave voltage that is a voltage required for the above.

  A preferable QAM signal generator used in the present invention includes a detector and a control device connected to a signal source for supplying a bias signal at each bias electrode. The control device supplies information on the bias signal to the signal source based on the information on the output light detected by the detector. In other words, the bias adjustment method can be automatically performed by providing this control device.

  Another preferred embodiment of the signal generator includes a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal provided in the first waveguide (2) in FIG. A second waveguide (5) having a coupling point (4) between the generator (3) and the first waveguide (2); and the second waveguide (5), A second QPSK signal generator (6) for outputting a signal combined with the output signal from the first QPSK signal generator; The present invention relates to a signal generation system having a multiplexing unit that combines outputs from a plurality of signal generators.

  That is, for example, by combining the outputs from the signal generators described above, it is possible to obtain optical signals with more variations. Therefore, as an example of “a signal combined with the output signal from the first QPSK signal generator”, a signal having half the amplitude of the output signal from the first QPSK signal generator can be cited. In this signal generation system, it is preferable that a signal is sent from one input section to each signal generation apparatus via a plurality of branch sections. In addition, as the signal generation system, the above-described configuration of the signal generation device can be adopted as appropriate.

  Another preferred embodiment of the signal generator includes a first waveguide (52) and a first four-phase phase shift keying (QPSK) signal generator (53) provided in the first waveguide (52). , A second waveguide (55) having a multiplexing point (54) with the first waveguide (53), and the first QPSK signal provided in the second waveguide (55). A second QPSK signal generator (56) for outputting a signal to be combined with an output signal from the generator, and a third waveguide having the second waveguide (55) and a combining point (57) And a third QPSK signal generator (59) provided in the third waveguide (58) and outputting a signal combined with the output signal from the second QPSK signal generator, , “N-th waveguide having a coupling point with the (n−1) -th waveguide, and the n-th waveguide, provided in the n-th waveguide, N-th QPSK signal generator "for outputting a signal combined with the output signal from the QPSK signal generator for all combinations of 4 or more and n or less, where n is an integer of 4 or more, It is a signal generator.

  That is, as shown in FIG. 15, various output signals can be obtained by providing a plurality of branches, but by providing a smaller QPSK signal generator, more various signals can be obtained. it can. By combining the outputs from the signal generators, it is possible to obtain a more varied optical signal. Therefore, as an example of “a signal combined with the output signal from the first QPSK signal generator”, a signal having half the amplitude of the output signal from the first QPSK signal generator can be cited. In this signal generation system, it is preferable that a signal is sent from one input section to each signal generation apparatus via a plurality of branch sections. Further, as the signal generator, the configuration of the signal generator described above can be adopted as appropriate.

  Another preferred embodiment of the signal generator includes a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal generator (3) provided in the first waveguide (2). , A second waveguide (5) having a multiplexing point (4) with the first waveguide (2), and the first QPSK signal provided in the second waveguide (5) The present invention relates to a signal generator comprising a second QPSK signal generator (6) for outputting a signal combined with an output signal from the generator.

  If the “signal combined with the output signal from the first QPSK signal generator” is a signal having half the amplitude of the output signal from the first QPSK signal generator, the QAM signal generation described above is performed. You can get the device. On the other hand, for example, even if the output from the first QPSK signal generator is the same as the output from the second QPSK signal generator, these signals are superimposed to obtain various modulation signals. In this signal generation system, it is preferable that a signal is sent from one input section to each signal generation apparatus via a plurality of branch sections. In addition, as the signal generation system, the above-described configuration of the signal generation device can be adopted as appropriate.

  Another preferred embodiment of the signal generator includes a first waveguide (52) and a first four-phase phase shift keying (QPSK) signal provided in the first waveguide (52) in FIG. A second waveguide (55) having a coupling point (54) between the generator (53) and the first waveguide (53); and the second waveguide (55), A combining point (57) between the second QPSK signal generator (56) for outputting a signal combined with the output signal from the first QPSK signal generator and the second waveguide (55) is provided. And a third QPSK signal that is provided in the third waveguide (58) and outputs a signal combined with the output signal from the second QPSK signal generator. And a generator (59).

  If the “signal combined with the output signal from the first QPSK signal generator” is a signal having half the amplitude of the output signal from the first QPSK signal generator, the QAM signal generation described above is performed. You can get the device. On the other hand, for example, even if the output from the first QPSK signal generator is the same as the output from the second QPSK signal generator, these signals are superimposed to obtain various modulation signals. In this signal generation system, it is preferable that a signal is sent from one input section to each signal generation apparatus via a plurality of branch sections. In addition, as the signal generation system, the above-described configuration of the signal generation device can be adopted as appropriate.

  As a QAM signal generation method, preferably, a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal generator (3) provided in the first waveguide (2). A second waveguide (5) having the first waveguide (2) and the multiplexing point (4), and a second QPSK signal generator provided in the second waveguide (5) (6), and the first QPSK signal generator (3) and the second QPSK signal generator (6) are respectively a main Mach-Zehnder waveguide (7a, 7b) and the main QPSK signal generator (6). A quadrature amplitude modulation (QAM) signal generator comprising two sub Mach-Zehnder waveguides (8a, 9a, 8b, 9b) provided on both arms of the Mach-Zehnder waveguides (7a, 7b) was used. A method of generating a QAM signal, wherein the first QQ The main Mach-Zehnder waveguide (7a) of the SK signal generator (3) and the main Mach-Zehnder waveguide (7b) of the second QPSK signal generator (6) each constitute a main Mach-Zehnder waveguide 2. The first bias electrode (10a, 10b) and the second bias electrode (11a, 11b) for adjusting the bias voltage applied to each of the two arms and the main Mach-Zehnder waveguide (7a, 7b) are combined. Including at least one bias electrode among third bias electrodes (13a, 13b) for applying a bias voltage to a signal propagating through both arms combined by the wave portions (12a, 12b), Four sub Mach-Zehnder waveguides (8a, 9a, 8b, 9b) propagate through both arms before the propagation signals are combined. The phase difference of the signal is 180 degrees, and each main Mach-Zehnder waveguide (7a, 7b) has a phase difference between the signals propagating through both arms before the propagation signals are combined, where m is an integer. (45 + 90 m) degrees, and the first QPSK signal generator (3) and the second QPSK signal generator (6) have the propagation signals before being combined at the multiplexing point (4). The first bias electrodes (10a, 10b), the first bias electrode (10a, 10b), the phase difference between signals propagating through the first waveguide (2) and the second waveguide (5) are 0 degrees or 180 degrees. The bias signal applied to the second bias electrode (11a, 11b) and the third bias electrode (13a, 13b) is controlled, and the output signal from the first QPSK signal generator and the second QPSK are controlled. Output signal from signal generator (6) And a method of generating a QAM signal, including a step of combining. In this QAM signal generation method, the configuration related to the QAM signal generation apparatus described above can be adopted as appropriate.

  As a method for generating the QAM signal, preferably, a first waveguide (2) and a first four-phase phase shift keying (QPSK) signal generator (3) provided in the first waveguide (2) are used. ) And a second waveguide (5) having a coupling point (4) between the first waveguide (2) and a second QPSK signal provided in the second waveguide (5) Generator (6), wherein the first QPSK signal generator (3) and the second QPSK signal generator (6) are respectively a main Mach-Zehnder waveguide (7a, 7b), Quadrature amplitude modulation (QAM) signal generator comprising two sub Mach-Zehnder waveguides (8a, 9a, 8b, 9b) provided on both arms of the main Mach-Zehnder waveguide (7a, 7b). A method of generating a QAM signal using The main Mach-Zehnder waveguide (7a) of the first QPSK signal generator (3) and the main Mach-Zehnder waveguide (7b) of the second QPSK signal generator (6) constitute a main Mach-Zehnder waveguide, respectively. A first bias electrode (10a, 10b) and a second bias electrode (11a, 11b) for adjusting a bias voltage applied to each of the two arms, and a main Mach-Zehnder waveguide (7a, 7b). And any one or more bias electrodes among the third bias electrodes (13a, 13b) for applying a bias voltage to a signal propagating through both arms combined in the multiplexing unit (12a, 12b). , The four sub-Mach-Zehnder waveguides (8a, 9a, 8b, 9b) have both arms before the propagation signals are combined. The phase difference of the transmitted signal is 180 degrees, and each main Mach-Zehnder waveguide (7a, 7b) has the phase difference of the signal propagating through both arms before the propagation signal is combined as l. The integer is (30 + 90l) degrees, and the first QPSK signal generator (3) and the second QPSK signal generator (6) have the propagation signals before being combined at the multiplexing point (4). , The first bias electrodes (10a, 10b), so that the phase difference between the signals propagating through the first waveguide (2) and the second waveguide (5) is 0 degree or 180 degrees, A bias signal applied to the second bias electrode (11a, 11b) and the third bias electrode (13a, 13b) is controlled, an output signal from the first QPSK signal generator, and the second signal From the QPSK signal generator (6) This is a method for generating a QAM signal, including a step of combining an output signal.

  Another preferable aspect of the present invention is that the main Mach-Zehnder waveguide (7a, 7b) and the two sub-Mach-Zehnder waveguides (8a) provided on both arms of the main Mach-Zehnder waveguide (7a, 7b) are provided. , 9a, 8b, 9b), a test signal applying device for applying a test signal to a bias electrode of either or both of the main Mach-Zehnder waveguide and the two sub-Mach-Zehnder waveguides, A light detection device for detecting an output signal from a waveguide to which a test signal is applied by a test signal application device, and an output spectrum detected by the light detection device using two detection data from the light detection device are two. If it contains sideband components, it has an automatic bias adjustment device that adjusts the bias voltage so that the double component of the modulation frequency becomes smaller A QPSK signal generator. In the QPSK signal generator, the configuration related to the QAM signal generator described above can be adopted as appropriate. In order to adjust the bias voltage so that the double component of the modulation frequency is reduced by the automatic bias adjustment device, the detection data from the light detection device may be monitored while increasing or decreasing the bias voltage.

  As the QPSK signal generator, preferably, the test signal is the apparatus according to any one of the above, whose frequency is 10 kHz or less.

  As the QPSK signal generator, preferably, the test signal is applied to both of the two sub Mach-Zehnder waveguides, and the two test signals have a phase difference of approximately 90 degrees or 270 degrees. Device.

  In the above description, the case where the optical QAM signal is used has been mainly described. However, as described above, the present invention is not limited to demodulating an optical QAM signal. For example, as an optical signal, any one of an optical QAM signal, an optical ASK signal, an optical PSK signal, an optical FSK signal, an optical CPFSK signal, an optical APSK signal, or an optical MSK signal that takes a value of 2 or more is appropriately used. it can.

  A case of demodulating an n-ASK signal (n-value ASK signal) will be described. In the constellation map of the nth order ASK signal, a peak appears along the I axis. Therefore, if a binary random number is given as a pseudo Q component to each received signal, a method similar to the method for demodulating the optical QAM signal described above can be used. After the signal identification, the n-ASK signal can be demodulated by using only the determination result for the I component.

  The optical PSK signal is a signal (face shift keying signal) having optical phase as modulation information. A case where a QPSK (Quadrature PSK) signal is demodulated will be described. In the constellation map of the QPSK signal, one peak appears in each quadrant. Therefore, the QPSK signal can be demodulated by using a method similar to the optical QAM demodulation method described above. A BPSK (binary PSK) signal can be demodulated in the same manner as the n-ASK signal described above. Even if the multilevel value of the PSK signal increases, the PSK signal can be demodulated by using the same method as the optical QAM demodulation method.

  In the constellation map of the optical APSK (intensity and phase shift keying) signal, peaks appear along the I axis and the Q axis. Therefore, the optical APSK signal can be demodulated by using the same method as the optical QAM demodulation method described above. Specifically, the optical APSK signal can be demodulated by using a boundary calculation step including a midpoint calculation step and a point group connection / extension step. Further, a demodulation method using a demodulator that further includes an apparatus for analyzing information related to frequency distribution such as a maximum likelihood sequence estimator (MLSE) is a preferred aspect of the present invention. The boundary shown in FIG. 6 is obtained by connecting a plurality of diagonal points obtained in the diagonal point calculation step and a plurality of adjacent midpoints obtained in the midpoint calculation step. On the other hand, in a preferred embodiment of the present invention, the boundary line is obtained using information on the frequency distribution. Specifically, the frequency of the optical QAM signal expressed in the histogram of the optical QAM signal based on the optical QAM signal on the constellation map is also taken into consideration so that the low frequency region becomes the boundary line. In this way, the optical APSK signal can be demodulated. In this case, the boundary line is not a diagonal point and a midpoint or a diagonal point connected as shown in FIG. For example, the intersection of the boundary line obtained as shown in FIG. 6 and a line connecting peaks that cross the boundary line is obtained. Then, the frequency distribution around the intersection is obtained, and the infrequent point is obtained. Then, connect the diagonal points and the infrequent points. In addition, the infrequent point is connected to the intersection of a line connecting peaks that cross the boundary line. In this way, a polygonal boundary line can be obtained. In this way, the optical APSK signal can be demodulated.

  In the case of an optical FSK signal, the position of each symbol changes with time on the constellation map. However, the speed of the phase change on the constellation map is equal to the frequency transition. Therefore, the signal is sampled a plurality of times at a sampling rate equal to or higher than the modulation rate, and the average value of the distances of the respective signals obtained by sampling is obtained. The average value obtained is the rate of phase change. The average value may be used for evaluation on the constellation map and demodulated. By performing such preprocessing, the optical FSK signal can also be demodulated in the same manner as the PSK signal.

  An optical CPFSK (Continuous Phase Frequency Shift Keying) signal is a kind of FSK signal. Therefore, the optical CPFSK signal can be demodulated by using a method similar to the method for demodulating the optical FSK signal. When demodulating the CPFSK signal, a demodulated signal by differential detection can also be obtained. For example, differential detection can be performed by using a signal before 1 Baud.

  The MSK (Minimum Shift Keying) signal is a kind of FSK signal. Therefore, the optical MSK signal can be demodulated by using a method similar to the method for demodulating the optical FSK signal.

Reference example 1
16QAM Mapping Using Two DPMZMs FIG. 16 shows a principle diagram of the 16QAM modulator of the present invention. In this example, a 16QAM signal is obtained by superimposing two QPSK signals having different amplitudes (intensities). The intensity difference between the two QPSK signals was 6 dB. A QPSK signal having a large amplitude (indicated by QPSK2 in FIG. 16) determines a region to which a quadrant is mapped. On the other hand, a QPSK signal having a small amplitude (indicated by QPSK1 in FIG. 16) fixes the position of each quaternary value. By combining two QPSK signals, 16 equally spaced symbols can be mapped in the phase diagram. This 16QAM mapping can be obtained by using binary data without using multi-level electric signals.

Reference Example 2 QPMZM for 16QAM
A LiNbO ₃ waveguide and a silicon-based PLC (planner lightwave circuit) are optically coupled to produce a modulator as shown in FIG. 17 having a QPMZM (four parallel MZMs) for 16QAM modulation. did. In this QPMZM, four MZMs (MZM-I, MZM-Q, MZM-i, and MZM-q) are coupled in parallel. In other words, this QPMZM has two DPMZMs (two parallel MZMs). Each MZM has traveling wave type electrodes (RFa1, RFb1, RFa2, and RFb2). As shown in FIG. 17, the QPMZM has six additional bias electrodes to control the phase offset between the MZMs. The two DPMZM inputs and outputs were coupled at their ends with a PLC-based optical coupler.

  The electro-optical response frequency response in each electrode is shown in FIG. In FIG. 18, the modulation electrodes having a bandwidth of 3 dB and 6 dB were about 10 GHz and 25 GHz, respectively. That is, it was found that this modulator can be used for high-speed modulation up to 25 Gbaud. The half-wave voltage of each MZM was 2.9 V for DC voltage and 4.2 V for 10 GHz. The insertion loss of the modulator used in this reference example was 10 dB.

50-Gb / s 16QAM Modulation and Demodulation FIG. 19 is a schematic configuration diagram of the apparatus used in this embodiment. On the transmitter side, 16QAM modulation was applied to continuous light from the external cavity semiconductor laser by QPMZM. Each arm of the modulator is connected to a binary NRZ (non-return-to-zero) PRBS with a data length of 2 ⁹ -1 and 12.5 Gb / s generated by a commercially available 4-channel pulse pattern generator. (pseudo random bit sequence) data was used for push-pull drive. A set of MZMs (MZM-I and MZM-Q) were driven in the range of -π to π so that a large amplitude QPSK signal was obtained. The remaining MZM pairs were driven in a range of −π / 2 to π / 2 so that a QPSK signal with a small amplitude was obtained.

  On the receiver side, the signal was demodulated using a digital homodyne detector. This digital homodyne detector uses an optical hybrid coupler to mix a signal and local oscillator (LO) light. For simplicity, LO and signal light are generated using a common semiconductor laser. The hybrid coupler gives a 90-degree phase offset between the four output ports, and the LO phase is set as a reference by differential detection of a set of [0 degree, 180 degrees] and [-90 degrees, 90 degrees]. The I component (in-phase component) and Q component (quadrature phase component) recovered. The detected optical signal was input to a high-speed AD converter. Then, the phase difference between the signal and the LO was calculated by a digital signal processor adjusted for the QAM signal. In this way, the I component and the Q component were recovered.

  FIG. 20A is a map (constellation map) obtained by driving the MZM-I and MZM-Q. FIG. 20B is a map obtained by driving MZM-i and MZM-q. 20A and 20B show that a QPSK signal having a large amplitude and a small amplitude can be generated, respectively. FIG. 20C is an IQ map obtained by driving all MZMs. It can be seen from FIG. 20C that an optical 16QAM signal has been generated.

The received I component and Q component are multilevel signals. This signal was decoded using the demodulator of the present invention that determines the signal threshold according to the distortion on the constellation map. Compared to the original data, the BER (bit error rate) is considered to be 2 × 10 ⁻³ , and demodulation almost equal to the upper limit of BER to which FEC (Forward Error Correction) is applicable was realized.

  From the obtained modulation spectrum shown in FIG. 21, it can be seen that the occupied band of 16QAM is the same as the band of DPSK or DQPSK. Thus, it was shown that a 50 Gb / s signal can be accommodated in a 12.5 Gb / s conventional WDM channel. With the use of polarization multiplexing, the two channels can be multiplexed to achieve a transfer of 100 GB / s. Furthermore, the frequency response of QPMZM was at a level sufficient to achieve 25 Gbaud, which was equivalent to 100 Gb / s 16QAM.

  As described above, the QPMZM of this embodiment can perform 16QAM by superimposing two QPSK. This example shows that 50 Gb / s optical 16QAM modulation can be achieved.

  FIG. 22 is a graph replaced with a drawing showing the distribution of the optical QAM signal and the boundary line.

Demodulation of 16-level optical QAM signal using DPMZM FIG. 23 is a schematic diagram of an apparatus used in the second embodiment. As shown in FIG. 23, in this embodiment, an optical SSB modulator is used to superimpose two 4-ASK signals to obtain a 16-level optical QAM signal. That is, two types of 2-bit information were prepared using an electrical signal and combined to obtain a 16-level optical QAM signal. The modulation rate was 2 GBaud (8 Gb / s).

FIG. 24 is a graph instead of a drawing showing a constellation map obtained in the second embodiment. The number of data plotted in FIG. 24 is 19799 points. FIG. 25 is a graph replaced with a drawing showing an optical QAM signal distribution and boundary lines in the second embodiment. The optical QAM signal was demodulated using the boundary line shown here. The BER of the I component was 1.6 × 10 ⁻³ , the BER of the Q component was 5.3 × 10 ⁻³ , and the overall BER was 3.4 × 10 ⁻³ .

  Since the present invention can provide a quadrature amplitude modulation signal generator, it can be suitably used in fields such as optical information communication.

FIG. 1 is a flowchart for explaining a signal demodulation method according to the present invention. FIG. 2 is a graph instead of a drawing showing an example of the optical QAM signal on the constellation map. FIG. 3 is a graph instead of a drawing showing an example of a histogram of an optical QAM signal based on the 16-value optical QAM signal on the constellation map of FIG. FIG. 4 is a diagram showing the peak positions and diagonal points of the histogram of the 16-value optical QAM signal on the constellation map. FIG. 5 is a diagram showing the peak positions based on the 16-value optical QAM signal on the constellation map and the midpoints of the outline symbols adjacent to each other. FIG. 6 is a diagram in which the boundary is obtained based on the 16-value optical QAM signal on the constellation map. FIG. 7A is a diagram illustrating an example of a constellation map of a QAM signal. FIG. 7B is a diagram showing a constellation map when preprocessing without weighting is performed on the signal of FIG. 7A. FIG. 7C is a diagram showing a constellation map when preprocessing for weighting with an exponential function is performed on the signal of FIG. 7A. FIG. 8 is a conceptual diagram showing a constellation map of the APSK signal. FIG. 9 is a conceptual diagram showing a constellation map after dummy symbols are installed. FIG. 10 is a diagram illustrating an example in which a boundary is obtained using only diagonal points belonging to the same quadrant. FIG. 11 is a diagram illustrating an example in which the boundary in the APSK signal is obtained. FIG. 12 is a diagram illustrating an example of processing for decoding a signal. FIG. 13 is a schematic diagram of a quadrature amplitude modulation signal generator according to the present invention. FIG. 14 is a diagram for explaining a quadrature amplitude modulation signal. FIG. 15 is a conceptual diagram of a QAM signal generator using three QPSK signal generators arranged in parallel. FIG. 16 shows a principle diagram of the 16QAM modulator of the present invention. FIG. 17 shows a schematic diagram of the apparatus in Reference Example 2. FIG. 18 shows the electro-optical response frequency response of each electrode in Reference Example 2. FIG. 19 is a schematic configuration diagram of the apparatus used in the first embodiment. FIG. 20A is a map (constellation map) obtained by driving the MZM-I and MZM-Q. FIG. 20B is a map obtained by driving MZM-i and MZM-q. 20A and 20B show that a QPSK signal having a large amplitude and a small amplitude can be generated, respectively. FIG. 20C is an IQ map obtained by driving all MZMs. FIG. 21 shows the spectrum of the optical signal obtained in Example 1. FIG. 22 is a graph replaced with a drawing showing the distribution of the optical QAM signal and the boundary line. FIG. 23 is a schematic diagram of the apparatus used in the second embodiment. FIG. 24 is a graph instead of a drawing showing a constellation map obtained in the second embodiment. FIG. 25 is a graph replaced with a drawing showing an optical QAM signal distribution and boundary lines in the second embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Quadrature amplitude modulation signal generator; 2 1st waveguide; 3 1st 4 phase shift keying signal generator; 4 Combined point; 5 2nd waveguide; 6 2nd QPSK signal generator

Claims (9)

A peak position calculating step for obtaining a peak position of a histogram of the received signal in the constellation map;
A boundary calculation step for obtaining a boundary for demodulating the signal based on the peak position obtained in the peak position calculation step;
A decoding step of decoding the received signal using the boundary obtained in the boundary calculation step , wherein the received signal takes a value of 2 or more, and an optical QAM signal, an optical ASK signal, an optical PSK; A signal, an optical FSK signal, an optical CPFSK signal, an optical APSK signal, or an optical MSK signal ,
The boundary calculation step includes:
A diagonal point calculation step for obtaining a diagonal point of a quadrangle formed by the four adjacent peaks, which is a peak obtained in the peak position calculation step;
A midpoint calculation step for obtaining a midpoint between adjacent peaks for the peaks obtained in the peak position calculation step and located in the outer contour;
A plurality of diagonal points obtained in the diagonal point calculating step and a plurality of middle points obtained in the middle point calculating step, connecting adjacent points, and a point group connecting step with a connecting line as a boundary; ,
Including ,
Demodulation method. The boundary calculation step further includes a step of providing a dummy symbol in the constellation map,
The step of providing a dummy symbol in the constellation map includes:
Calculating the inner product of two peaks in the constellation map,
Determining whether the inner product value obtained in the step of obtaining the inner product is within a predetermined range; and
When the inner product value is within a predetermined range, obtaining a sum vector of two peaks for which the inner product is obtained, and installing a dummy symbol at a position corresponding to the sum vector ;
Is a process of repeatedly performing
In the boundary calculation step, a boundary is obtained based on the position of the dummy symbol and the peak position obtained in the peak position calculation step.
The demodulation method according to claim 1. The peak position calculating step includes a preprocessing step,
The pretreatment step is
Including an amplitude correction step, a weighting step, a fourth power calculation step regarding the received signal, an average calculation step of the fourth power value, and a phase shift step,
The amplitude correction step is a step of correcting the amplitude of the in-phase component corresponding to the peak position in the constellation map and correcting the amplitude of the quadrature component,
The weighting step is a step of weighting the corrected in-phase component and quadrature component,
The fourth power calculation step related to the received signal is a step of obtaining a value related to phase drift by performing a fourth power operation on the weighted peak value,
The average calculation step of the fourth power value is a step of obtaining a time average value of values related to the phase drift,
The phase shift step is a step of obtaining a phase shift value based on the time average value, shifting the phase of the received signal by a predetermined amount based on the phase shift value, and obtaining a peak after phase shift correction,
The boundary calculation step is a step of obtaining a boundary based on the peak position after the phase shift correction.
The demodulation method according to claim 1 . The received signal weighting step includes:
This is a process to weight each in-phase component and quadrature component of the received signal used to obtain each peak using an exponential function
The demodulation method according to claim 3 . The received signal weighting step includes:
It is a process of weighting these received signals by extracting the received signals located at the four corners of the constellation map.
The demodulation method according to claim 3 . The received signal weighting step includes:
Extracting peaks located at the four corners of the constellation map;
Extracting a peak located in the center of the constellation map;
Weighting so that the peaks located at the four corners of the extracted constellation map become larger;
Weighting so that the peak located at the center of the extracted constellation map becomes larger;
including,
The demodulation method according to claim 3 . The boundary line is a curve
The demodulation method according to claim 1 . A peak position calculating means for obtaining a peak position of a histogram of the received signal in the constellation map;
Boundary calculating means for obtaining a boundary for demodulating a signal based on the peak position obtained by the peak position calculating means;
Decoding means for decoding a received signal using the boundary obtained by the boundary calculation means , wherein the received signal takes a value of 2 or more, and an optical QAM signal, an optical ASK signal, an optical PSK signal , An optical FSK signal, an optical CPFSK signal, an optical APSK signal, or an optical MSK signal ,
The boundary calculation means includes
Diagonal point calculation means for obtaining a diagonal point of a quadrangle formed by the four adjacent peaks, which is a peak obtained by the peak position calculation means;
Regarding the peak obtained by the peak position calculating means and located at the outer contour,
A midpoint calculation means for obtaining a midpoint between adjacent peaks;
A plurality of diagonal points obtained by the diagonal point calculating means and a plurality of midpoints obtained by the middle point calculating means, connecting adjacent ones, and a point group connecting means having a connecting line as a boundary; ,
Including signal demodulator. The boundary calculation means further includes means for providing a dummy symbol in the constellation map,
Means for providing dummy symbols in the constellation map are:
Calculating the inner product of two peaks in the constellation map,
Determining whether the inner product value obtained in the step of obtaining the inner product is within a predetermined range; and
When the inner product value is within a predetermined range, obtaining a sum vector of two peaks for which the inner product is obtained, and installing a dummy symbol at a position corresponding to the sum vector ;
Is a means of repeatedly performing
The boundary calculating means obtains a boundary based on the position of the dummy symbol and the peak position obtained by the peak position calculating means;
The demodulation method according to claim 8.

Images