JP5620246B2 - OSNR calculation method and apparatus - Google Patents

Description

  The present invention relates to an optical transmission technique, and more particularly to an OSNR calculation technique for calculating an OSNR (Optical Signal-to-Noise Ratio) indicating an intensity ratio between a signal and noise included in signal light.

  In order to economically realize large-capacity transmission, a transparent WDM (Wavelength Division Multiplex) optical network that does not involve photoelectric conversion at a node is being introduced. In a transparent optical network, it is important to monitor the optical signal quality and know the state of the network in order to establish connection and identify faulty sections. Among optical signal qualities, OSNR (Optical Signal-to-Noise Ratio) is an important parameter.

  FIG. 17 is an explanatory diagram showing a conventional OSNR calculation method. Conventionally, as shown in FIG. 17, the OSNR is calculated by an ASE (Amplified Spontaneous Emission) interpolation method that estimates the noise level on the frequency grid from the noise level at the end of the transmission band in the spectrum acquired by the OSA (Optical Spectrum Analyzer). (For example, refer nonpatent literature 1 etc.).

IEC 61280-2-9 Edition 2.0

  However, when the optical transmission speed is increased, the signal spectrum spreads and overlaps with the signal spectrum of the adjacent channel. Further, the transmission band is generated by the optical filter effect such as WSS (Wavelength Selective Switch) arranged in the transmission path. Edge noise is cut off. Therefore, the conventional ASE complement method has a problem that it is difficult to accurately calculate the OSNR.

  FIG. 18 is an explanatory diagram showing an overlap with a signal spectrum of an adjacent channel. Here, as compared with FIG. 17 described above, the signal spectrum of the optical signal is broadened as a whole, and an overlap occurs between the signal spectrum of the adjacent channel and the bottom of the signal spectrum. For this reason, an error occurs in the noise intensity measured at the frequency with the lowest intensity between the signal spectra.

  FIG. 19 is an explanatory diagram showing the blocking of noise at the end of the transmission band. Here, in comparison with FIG. 17 described above, the intensity between signal spectra is reduced to a level lower than the original noise intensity by the optical filter effect such as WSS arranged in the transmission path. For this reason, an error occurs in the noise intensity measured at the frequency with the lowest intensity between the signal spectra.

  The present invention has been made to solve such problems, and an object of the present invention is to provide an OSNR calculation technique capable of calculating an OSNR with little error even when the optical transmission speed is high.

In order to achieve such an object, an OSNR calculation method according to the present invention includes an autocorrelation function acquisition step of acquiring an autocorrelation function of a time waveform related to signal light from input signal light, and a peak of the autocorrelation function. The value is the sum of the signal strength and noise strength, the signal strength is estimated based on the autocorrelation function other than the peak value (autocorrelation value), and the signal strength is subtracted from the sum. And an OSNR calculation step of calculating an OSNR of the signal light from the signal intensity and the noise intensity . In this OSNR calculation step, the correlation value at a set time position different from the peak value in the autocorrelation function is calculated by the OSNR. Divide by a constant indicating the ratio between the autocorrelation value at the set time position and the signal intensity of the reference light, which is obtained in advance from the known autocorrelation function of the reference light By, in which to calculate the signal intensity.

In addition, another OSNR calculation method according to the present invention includes an autocorrelation function acquisition step of acquiring an autocorrelation function of a time waveform related to signal light from input signal light, and a peak value of the autocorrelation function as signal intensity and noise. The signal strength is calculated by estimating the signal strength based on the strength value other than the peak value (autocorrelation value) in the autocorrelation function and subtracting the signal strength from the strength sum. And an OSNR calculation step for calculating the OSNR of the signal light from the noise intensity. In this OSNR calculation step, a plurality of self values that are different from the peak value extracted from the autocorrelation function and that do not sandwich the peak value therebetween. from the correlation value is obtained by so specifying the approximate line of the autocorrelation function, and the signal strength values in the time position of the peak value of the approximate line.

  In addition, in the autocorrelation function acquisition step, a frequency spectrum measurement step for measuring the frequency spectrum of the signal light, and a part or all of any one pass band from the frequency spectrum is cut out as a finite section, There may be provided an autocorrelation function calculating step for calculating an autocorrelation function by performing inverse Fourier transform on the corrected spectrum to which the temporary intensity value is added outside.

  In addition, the autocorrelation function acquisition step is multiplied by a frequency spectrum measurement step for measuring the frequency spectrum of the signal light, and the frequency spectrum is multiplied by a window function having any one passband as a finite interval, and then inverse Fourier transform is performed. An autocorrelation function calculating step for calculating an autocorrelation function by performing a deconvolution operation of the window function on the data obtained in this manner may be provided.

  Also, in the autocorrelation function acquisition step, a frequency spectrum measurement step for measuring the frequency spectrum of the signal light, and for each inclination selected from the search range, the frequency spectrum is multiplied or divided by a linear function having the inclination. After correcting the slope of the frequency spectrum, an inverse Fourier transform is performed to provide an autocorrelation function calculation step for calculating an autocorrelation function for each slope. In the OSNR calculation step, the OSNR is calculated for each autocorrelation function for each slope. Of these OSNRs, the OSNR indicating the extreme value may be used as the OSNR of the signal light.

In addition, the OSNR calculation apparatus according to the present invention includes an autocorrelation function acquisition unit that acquires an autocorrelation function of a time waveform related to signal light from input signal light, a peak value of the autocorrelation function, and a signal intensity and a noise intensity. The signal strength is estimated based on the intensity value other than the peak value (autocorrelation value) of the autocorrelation function, and the noise intensity is obtained by subtracting the signal intensity from the intensity sum. An OSNR calculation unit for calculating the OSNR of the signal light from the intensity, and the correlation value at a set time position different from the peak value in the autocorrelation function is obtained from the autocorrelation function of the reference light whose OSNR is known. The signal strength is calculated by dividing by a constant indicating the ratio between the autocorrelation value at the set time position and the signal strength of the reference light, which has been obtained in advance. That.

In addition, another OSNR calculation apparatus according to the present invention includes an autocorrelation function acquisition unit that acquires an autocorrelation function of a time waveform related to signal light from input signal light, a peak value of the autocorrelation function as signal intensity and noise. The signal strength is calculated by estimating the signal strength based on the strength value other than the peak value (autocorrelation value) in the autocorrelation function and subtracting the signal strength from the strength sum. And an OSNR calculation unit for calculating the OSNR of the signal light from the noise intensity, wherein the OSNR calculation unit extracts a plurality of autocorrelations that are different from the peak values that are extracted from the autocorrelation function and that do not sandwich the peak values. from the value identifies the approximate line of the autocorrelation function is obtained by such a signal strength value in the time position of the peak value of the approximate line.

  In addition, the autocorrelation function acquisition unit extracts a part or all of one of the pass bands from the frequency spectrum measurement unit that measures the frequency spectrum of the signal light and the frequency spectrum as a finite interval, An autocorrelation function calculating unit that calculates an autocorrelation function may be provided by performing inverse Fourier transform on the corrected spectrum to which a temporary intensity value is added outside.

  According to the present invention, the optical transmission speed is increased, the signal spectrum spreads and overlaps with the signal spectrum of the adjacent channel, and further, the light such as WSS (Wavelength Selective Switch) arranged in the transmission path is generated. Even when the noise at the end of the transmission band is blocked by the filter effect, the OSNR with less error can be calculated.

It is a block diagram which shows the structure of the OSNR calculation apparatus concerning 1st Embodiment. It is an example of an autocorrelation function of a signal time waveform. It is an example of an autocorrelation function of a noise time waveform. It is an example of an autocorrelation function of a time waveform obtained from signal light including a signal and noise. It is a flowchart which shows the OSNR calculation process concerning 1st Embodiment. It is explanatory drawing which shows the example of an estimation of the signal strength concerning 2nd Embodiment. It is explanatory drawing which shows the example of an estimation of the signal strength concerning 3rd Embodiment. It is a block diagram which shows the structure of the OSNR calculation apparatus concerning 4th Embodiment. It is explanatory drawing which shows the relationship between the space | interval of an autocorrelation value, and a base width. It is explanatory drawing of a correction spectrum. It is explanatory drawing which shows the change of OSNR by the inclination of a frequency spectrum. It is explanatory drawing which shows the example of the inverse Fourier transform using a linear function. It is explanatory drawing which shows the autocorrelation function of a time waveform in case a frequency spectrum has inclination. It is a flowchart which shows the OSNR calculation process concerning 7th Embodiment. It is a block diagram which shows the structure of the OSNR calculation apparatus 10 concerning 8th Embodiment. It is a block diagram which shows the structure of an autocorrelation function measurement part. It is explanatory drawing which shows the conventional OSNR calculation method. It is explanatory drawing which shows the overlap with the signal spectrum of an adjacent channel. It is explanatory drawing which shows the interruption | blocking of the noise in a transmission band edge.

Next, embodiments of the present invention will be described with reference to the drawings.
[First Embodiment]
First, the OSNR calculation apparatus according to the first embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram illustrating a configuration of the OSNR calculation apparatus according to the first embodiment.

  The OSNR calculation device 10 is a device that calculates an OSNR (Optical Signal-to-Noise Ratio) indicating an intensity ratio between a signal and noise included in the input signal light L. The OSNR calculation apparatus 10 includes an autocorrelation function acquisition unit 11 and an OSNR calculation unit 12 as main functional units.

The autocorrelation function acquisition unit 11 has a function of acquiring an autocorrelation function of a time waveform from the input signal light L.
The OSNR calculation unit 12 has a function of estimating the signal intensity and the noise intensity from the autocorrelation function of the time waveform acquired by the autocorrelation function acquisition unit 11 and calculating the OSNR of the signal light L from the signal intensity and the noise intensity. doing.

[Principle of the Invention]
The principle of the present invention will be described with reference to FIGS. FIG. 2 is an example of an autocorrelation function of a signal time waveform. FIG. 3 is an example of an autocorrelation function of a noise time waveform. FIG. 4 is an example of an autocorrelation function of a time waveform obtained from signal light including a signal and noise.

  In the present invention, the OSNR is calculated by utilizing the fact that the autocorrelation functions in the signal and noise time waveforms are different. FIG. 2 shows, as an example, an autocorrelation function of an ideal time waveform related to a QPSK-modulated signal included in signal light. This autocorrelation function has a triangular wave shape as a whole, has a peak value (maximum value) at time T0, and its intensity corresponds to the signal intensity S.

  On the other hand, FIG. 3 shows, as an example, an autocorrelation function of a time waveform related to noise included in signal light. The autocorrelation function in FIG. 3 is a delta function. This autocorrelation function has a peak value (maximum value) at time T0, and its intensity corresponds to the noise intensity N.

Since the signal light received via the transmission line includes such a signal and noise, the autocorrelation function of the time waveform is the autocorrelation function of the time waveform related to the signal and noise as shown in FIG. Becomes a synthesized waveform.
Here, the peak value of the autocorrelation function in FIG. 4 corresponds to the sum of the peak values of the autocorrelation function related to the signal and noise, that is, the sum S + N of the signal strength S and the noise strength N. Therefore, since the noise intensity N of the autocorrelation function related to noise is synthesized at the time T0 corresponding to the signal intensity S of the autocorrelation function related to the signal, only the signal intensity S or the noise intensity N cannot be acquired. .

  The present invention pays attention to the fact that the signal and noise modulated in this way are significantly different in the waveform of the autocorrelation function of the time waveform and can be separated, and among the autocorrelation functions of the time waveform acquired from the signal light, The signal intensity S is estimated based on the autocorrelation function part of the time waveform related to the signal, that is, the intensity value (autocorrelation value) other than the peak value.

Then, the noise intensity N is calculated by subtracting the signal intensity S from the intensity sum S + N, and the ratio of the signal intensity S and the noise intensity N is calculated to calculate the OSNR of the input signal light. It is.
In the present invention, a QPSK modulation signal will be described as an example. However, since autocorrelation functions of an ideal signal and a noise time waveform are generally different in shape, the same effect can be expected in other modulation formats.

[OSNR calculation device]
Next, the configuration of the OSNR calculation apparatus 10 according to the present embodiment will be described in detail with reference to FIG.
The OSNR calculation device 10 is composed of an arithmetic processing device such as a server device as a whole, and an autocorrelation function acquisition unit 11 and an OSNR calculation unit 12 are provided as main functional units. In addition, the OSNR calculation device 10 is provided with functions provided in an arithmetic processing device such as a general server device such as a data communication function, an operation input function, a screen display function, and a data storage function. .

The autocorrelation function acquisition unit 11 has a function of acquiring an autocorrelation function of a time waveform related to the signal light L from the input signal light L.
The OSNR calculation unit 12 has a function of setting the peak value of the autocorrelation function obtained by the autocorrelation function obtaining unit 11 to the sum S + N of the signal intensity and the noise intensity, and the autocorrelation function obtained by the autocorrelation function obtaining unit 11. A function of estimating the signal intensity based on the intensity value relating to the signal, a function of obtaining the noise intensity N by subtracting the signal intensity S from the intensity sum S + N, and the signal light L from the signal intensity S and the noise intensity N And a function of calculating OSNR.

[Operation of First Embodiment]
Next, the operation of the OSNR calculation apparatus 10 according to the present embodiment will be described with reference to FIG. FIG. 5 is a flowchart illustrating OSNR calculation processing according to the first embodiment.
The OSNR calculation device 10 executes the OSNR calculation process of FIG. 5 in response to a process start instruction from the operator or an input of the signal light L.

First, the autocorrelation function acquisition unit 11 acquires an autocorrelation function of a time waveform related to the signal light L from the input signal light L (step 100).
Next, the OSNR calculation unit 12 calculates the peak value of the autocorrelation function obtained by the autocorrelation function acquisition unit 11 as the sum S + N of the signal intensity and the noise intensity (step 101), and the autocorrelation function acquisition unit 11 The signal strength is estimated based on the strength value related to the signal among the obtained autocorrelation functions (step 102).

Thereafter, the OSNR calculation unit 12 calculates the noise intensity N by subtracting the signal intensity S from the intensity sum S + N (step 103), and based on the signal intensity S and the noise intensity N based on the following equation (1). Then, the OSNR of the signal light L is calculated (step 104).

[Effect of the first embodiment]
Thus, in the present embodiment, the autocorrelation function acquisition unit 11 acquires the autocorrelation function of the time waveform related to the signal light from the input signal light L, and the OSNR calculation unit 12 acquires the autocorrelation function of the autocorrelation function. By making the peak value the intensity sum S + N of the signal intensity S and the noise intensity N, estimating the signal intensity S based on the intensity value related to the signal in the autocorrelation function, and subtracting the signal intensity S from the intensity sum S + N The noise intensity N is obtained, and the OSNR of the signal light L is calculated from the signal intensity S and the noise intensity N.

  As a result, the optical transmission speed is increased, and the spread of the signal spectrum causes an overlap with the signal spectrum of the adjacent channel. Furthermore, due to the optical filter effect such as WSS (Wavelength Selective Switch) arranged in the transmission path. Even when the noise at the end of the transmission band is cut off, the OSNR with less error can be calculated.

[Second Embodiment]
Next, an OSNR calculation apparatus 10 according to the second embodiment of the present invention will be described with reference to FIG. FIG. 6 is an explanatory diagram of an example of signal strength estimation according to the second embodiment.

In the present embodiment, a specific configuration of the OSNR calculation unit 12 will be described.
In the present embodiment, the OSNR calculation unit 12 uses a correlation value at a set time position different from the peak value among the autocorrelation functions of the time waveform related to the signal light obtained by the autocorrelation function acquisition unit 11, and the OSNR is known. It has a function to calculate the signal intensity by dividing by a constant α, which is obtained in advance from the autocorrelation function of the reference light and is a ratio α between the autocorrelation value at the set time position and the signal intensity of the reference light. ing.

  As shown in FIG. 6, the autocorrelation function of the time waveform related to the signal included in the signal light L has a triangular wave shape, and the vertex thereof corresponds to the signal intensity S. Actually, since the two autocorrelation functions related to the signal and noise are combined at the time T0 which is the time position of the peak value, it is impossible to acquire only the signal intensity S itself from the autocorrelation function.

  Here, since the autocorrelation function relating to the signal has a triangular wave shape, it can be considered that the time and the autocorrelation value (intensity) are in a proportional relationship on the hypotenuse. Therefore, if an approximate line that approximates the hypotenuse is obtained, the peak value at time T 0, that is, the signal intensity S can be estimated based on the autocorrelation value at the set time position.

  In general, the shape of an autocorrelation function of a time waveform related to a signal depends on optical transmission conditions such as a multiplexing communication system and transmission characteristics of a transmission line. For this reason, if the autocorrelation function of the time waveform is acquired in advance for the reference light having a known OSNR obtained by transmission under the optical transmission conditions, the ratio between the time and the autocorrelation value for the hypotenuse, that is, The slope of the hypotenuse can be obtained.

  In the present embodiment, as shown in FIG. 6, the autocorrelation value at the set time position T1 separated from the time T0 by the unit time t is α times the signal intensity S corresponding to the apex (0 <α <1). If so, the slope of the hypotenuse is more specifically defined. This constant α corresponds to the reciprocal of the slope related to the hypotenuse of the autocorrelation function obtained from the reference light.

Therefore, in step 102 in FIG. 5, the OSNR calculation unit 12 has an autocorrelation value different from the peak value out of the autocorrelation function of the time waveform related to the input signal light L, and from the peak value time T0. The signal intensity S can be obtained by obtaining the autocorrelation value C1 = αS at the set time position T1 separated by the unit time T and dividing the autocorrelation value C1 by the constant α obtained in advance from the reference light. it can. Thus, when the autocorrelation value at time T0 is C0 (= S + N), the OSNR of the signal light L is calculated by the following equation (2).

[Effect of the second embodiment]
As described above, in the present embodiment, in the OSNR calculation unit 12, the correlation value at the set time position T1 different from the peak value among the autocorrelation functions of the time waveform related to the signal light obtained by the autocorrelation function acquisition unit 11. The signal intensity S is calculated by dividing C1 by a constant α obtained in advance from the autocorrelation function of the reference light whose OSNR is known.
Thereby, the calculation for calculating the OSNR becomes very simple. For example, when this calculation is implemented in hardware such as a measuring instrument, hardware resources such as memory can be saved. Further, the calculation time can be shortened, and an effect that the OSNR can be calculated at high speed is obtained.

  In the present embodiment, the case where the autocorrelation function related to the signal is assumed to have a triangular wave shape has been described. However, the present invention is not limited to this. For example, unlike the triangular wave shape, when the autocorrelation function of the time waveform of the signal is distorted due to the transmission path characteristics or the like, from the autocorrelation function of the time waveform including the transmission path characteristics whose OSNR is known, By obtaining the constant α, the OSNR can be calculated.

[Third Embodiment]
Next, an OSNR calculation apparatus 10 according to the third embodiment of the present invention will be described with reference to FIG. FIG. 7 is an explanatory diagram of an example of signal strength estimation according to the third embodiment.

  In the second embodiment, the case where the signal intensity S is estimated using the constant α obtained from the reference light whose OSNR is known has been described as an example. In the present embodiment, an approximate line of the autocorrelation function is generated from a plurality of autocorrelation values extracted from the autocorrelation function of the time waveform related to the signal light obtained by the autocorrelation function acquisition unit 11, and this approximate line is generated. A case where the signal intensity is estimated from the above will be described.

  In the present embodiment, the OSNR calculation unit 12 is different from the peak value among the autocorrelation functions of the time waveform related to the signal light obtained by the autocorrelation function acquisition unit 11, and does not sandwich the peak value. A function of extracting a plurality of autocorrelation values, a function of specifying an approximate line FR of the autocorrelation function from these extracted autocorrelation values, and a signal at the time position T0 of the peak value of the obtained approximate line FR And a function for estimating the intensity S.

  As shown in FIG. 7, the autocorrelation function of the time waveform related to the signal included in the signal light L has a triangular wave shape, and the vertex thereof corresponds to the signal intensity S. Actually, since the two autocorrelation functions related to the signal and noise are combined at the time T0 which is the time position of the peak value, it is impossible to acquire only the signal intensity S itself from the autocorrelation function.

  Here, since the autocorrelation function relating to the signal has a triangular wave shape, the approximate line FR relating to the hypotenuse can be identified as a regression line from a plurality of autocorrelation values (intensities) on the hypotenuse. At this time, the approximate line FR only needs to approximate either the left or right hypotenuse as viewed from the time T0. For this reason, the autocorrelation value used for specifying the approximate line FR may be extracted so as not to sandwich the peak value.

  In FIG. 7, two autocorrelation values C1 and C2 are extracted. Therefore, the functional equation of the approximate line FR can be specified from these autocorrelation values C1 and C2 and the time interval between these autocorrelation values C1 and C2. Note that the interval between the autocorrelation values is determined when the autocorrelation function of the time waveform related to the signal light L is acquired. For example, when the bandwidth of the frequency spectrum of the signal light L is B, the autocorrelation value interval is 1 / B on the autocorrelation function of the time waveform obtained by inverse Fourier transform of the frequency spectrum.

Therefore, in step 102 in FIG. 5, the OSNR calculation unit 12 determines a plurality of self-correlation functions that are different from the peak value and do not sandwich the peak value among the autocorrelation functions of the time waveform related to the input signal light L. Correlation values C1 and C2 are extracted, and an approximate line FR of the autocorrelation function is specified from these autocorrelation values by a regression line using interpolation, extrapolation, least squares, or the like, and the obtained approximate line FR is obtained. Thus, the signal intensity S can be obtained by interpolating the value of the peak value at time T0. Thus, when the autocorrelation value at time T0 is C0 (= S + N), the OSNR of the signal light L is calculated by the following equation (3).

[Effect of the third embodiment]
As described above, in the present embodiment, the OSNR calculation unit 12 differs from the peak value out of the autocorrelation functions of the time waveform related to the signal light obtained by the autocorrelation function acquisition unit 11 and calculates the peak value. A plurality of autocorrelation values that are not sandwiched between them are extracted, an approximate line FR of the autocorrelation function is specified from these autocorrelation values, and the value at the time position T0 of the peak value of the approximate line FR is estimated as the signal intensity S It is what you do.
This eliminates the need to obtain the constant α in advance using the reference light as in the second embodiment, and avoids the preprocessing burden required for OSNR calculation.

[Fourth Embodiment]
Next, with reference to FIG. 8, an OSNR calculation apparatus 10 according to a fourth embodiment of the present invention will be described. FIG. 8 is a block diagram showing the configuration of the OSNR calculation apparatus according to the fourth embodiment, and the same or equivalent parts as those in FIG. 1 are given the same reference numerals.

In the present embodiment, a specific configuration of the autocorrelation function acquisition unit 11 will be described.
In the present embodiment, the autocorrelation function acquisition unit 11 is provided with a spectrum measurement unit 11A and an autocorrelation function calculation unit 11B as main functional units.

  The spectrum measuring unit 11A has a function of measuring the frequency spectrum of the input signal light L. The spectrum measuring unit 11A can be configured with an optical filter, a dispersion spectrometer, a Fourier transform spectrometer, a Hadamard transform spectrometer, and the like.

  The autocorrelation function calculation unit 11B has a function of cutting out a part or all of any one pass band from the frequency spectrum obtained by the spectrum measurement unit 11A as a finite interval, and a finite interval of the obtained partial spectrum. It has a function of generating a corrected spectrum in which a temporary intensity value is added outside, and a function of calculating an autocorrelation function by performing inverse Fourier transform on the obtained corrected spectrum.

In general, the autocorrelation function of a signal's time waveform is obtained by performing an inverse Fourier transform by extracting a part or all of the passband included in the signal's frequency spectrum as a partial spectrum, for example, by applying a window function. Can do.
In this way, when the autocorrelation function is obtained by inverse Fourier transform, when the bandwidth of the partial spectrum is B, the interval of the autocorrelation values in the autocorrelation function of the time waveform obtained by inverse Fourier transform is 1 / B. Become.

  On the other hand, when the symbol rate of the signal is fs, the width of the low side of the triangular waveform after the inverse Fourier transform is 2 / fs. FIG. 9 is an explanatory diagram showing the relationship between the autocorrelation value interval and the base width. For this reason, if a partial spectrum having a sufficiently wide band satisfying 1 / B <1 / fs is not used, points representing a triangular waveform are lost.

  In the present embodiment, in step 101 of FIG. 5, the spectrum measurement unit 11A measures the frequency spectrum from the signal light L, and the autocorrelation function calculation unit 11B cuts out from this frequency spectrum and obtains the partial spectrum obtained. An autocorrelation function is calculated by adding a temporary intensity value (0 point) D outside the effective interval to generate a corrected spectrum and performing inverse Fourier transform on the corrected spectrum.

  FIG. 10 is an explanatory diagram of the corrected spectrum. Here, the bandwidth B of the partial spectrum is an effective section, and a temporary intensity value D is added outside the effective section, so that the bandwidth B is expanded to B ′ in a pseudo manner. Accordingly, since the interval between the autocorrelation values can be reduced to 1 / B ′ (<1 / B), more points can be provided on the triangular waveform.

[Effect of the fourth embodiment]
Thus, in this embodiment, the frequency spectrum measurement unit 11A of the autocorrelation function acquisition unit 11 measures the frequency spectrum of the input signal light L, and the autocorrelation function calculation unit 11B A partial spectrum with a part or all of one of the passbands as a finite interval is cut out, and a corrected spectrum with a temporary intensity value added outside the finite interval is generated, and this corrected spectrum is subjected to inverse Fourier transform. By doing so, an autocorrelation function is calculated.
As a result, since the bandwidth of the partial spectrum can be broadened from B to B ′ in a pseudo manner, the OSNR can be calculated with high accuracy from the frequency spectrum with a narrow bandwidth, which is a case where 1 / B> 1 / fs. Can do.

  In the present embodiment, when the provisional intensity value D is added to the partial spectrum, an example in which the provisional intensity value D is added symmetrically about the center frequency f0 of the passband has been described. It is not limited to. For example, the same effect can be obtained even if the provisional intensity value D is added asymmetrically to the left or right, such as only one of the left and right.

[Fifth Embodiment]
Next, an OSNR calculation apparatus 10 according to the fifth embodiment of the present invention will be described.
In the fourth embodiment, the case where the autocorrelation function of the time waveform is calculated by performing inverse Fourier transform on the frequency spectrum of the signal light has been described as an example. In the present embodiment, distortion compensation of an autocorrelation function in the case where an autocorrelation function is calculated by performing inverse Fourier transform on a frequency spectrum will be described.

In the present embodiment, the autocorrelation function acquisition unit 11 is provided with a spectrum measurement unit 11A and an autocorrelation function calculation unit 11B as main functional units, as in FIG. 8 described above.
The spectrum measuring unit 11A has a function of measuring the frequency spectrum of the input signal light L.
The autocorrelation function calculation unit 11B multiplies the frequency spectrum obtained by the spectrum measurement unit 11A by a convex window function so that a part or all of any one pass band is limited to a finite interval from the frequency spectrum. And a function of calculating an autocorrelation function by performing inverse Fourier transform on the obtained partial spectrum.

  When calculating the autocorrelation function of the time waveform by performing inverse Fourier transform on the frequency spectrum of the signal light, the autocorrelation function calculating unit 11B extracts a partial spectrum from the frequency spectrum using a window function. Become. At this time, when clipping is performed using a rectangular window function, distortion is generated in the autocorrelation function obtained by inverse Fourier transform due to the influence of the processing.

In the present embodiment, in step 101 of FIG. 5, the spectrum measurement unit 11A measures the frequency spectrum from the signal light L, and the autocorrelation function calculation unit 11B multiplies the frequency spectrum by a convex window function. The autocorrelation function is calculated by cutting out an arbitrary bandwidth as an effective section and subjecting the obtained partial spectrum to inverse Fourier transform.
There are several convex window functions. By using window functions such as Han, Hamming, and Blackman, it was confirmed that the OSNR error was reduced due to autocorrelation function distortion.

[Effect of Fifth Embodiment]
As described above, in the present embodiment, the autocorrelation function calculation unit 11B multiplies the frequency spectrum obtained by the spectrum measurement unit 11A by the convex window function to thereby select any one of the frequency spectra. A part or all of the pass band is cut out as a finite section, and the autocorrelation function is calculated by performing inverse Fourier transform on the obtained partial spectrum.
As a result, the distortion generated in the autocorrelation function obtained by the inverse Fourier transform can be suppressed under the influence of the clipping process by the window function, and the OSNR error can be reduced.

[Sixth Embodiment]
Next, an OSNR calculation apparatus 10 according to the sixth embodiment of the present invention will be described.
In the case where the autocorrelation function is calculated by performing inverse Fourier transform on the frequency spectrum, the case where the distortion of the autocorrelation function is compensated by cutting out the partial spectrum using a convex window function has been described as an example. In this embodiment, a case will be described in which distortion of an autocorrelation function is compensated by performing a deconvolution operation with a window function on an autocorrelation function obtained by inverse Fourier transform.

In the present embodiment, the autocorrelation function acquisition unit 11 is provided with a spectrum measurement unit 11A and an autocorrelation function calculation unit 11B as main functional units, as in FIG. 8 described above.
The spectrum measuring unit 11A has a function of measuring the frequency spectrum of the input signal light L.
The autocorrelation function calculation unit 11B is a function of cutting out part or all of any one pass band as a finite section from the frequency spectrum by multiplying the frequency spectrum obtained by the spectrum measurement unit 11A by a window function. And a function of calculating an autocorrelation function by performing inverse Fourier transform on the obtained partial spectrum, and a function of performing a deconvolution operation on the obtained autocorrelation function using a window function.

  As described above, when the autocorrelation function of the time waveform is calculated by performing inverse Fourier transform on the frequency spectrum of the signal light, the autocorrelation function calculation unit 11B partially converts the data of a certain bandwidth using the window function from the frequency spectrum. The spectrum will be cut out. At this time, when clipping is performed using a window function, distortion occurs in the autocorrelation function obtained by inverse Fourier transform due to the influence of the processing.

  In the present embodiment, in step 101 of FIG. 5, the spectrum measurement unit 11A measures the frequency spectrum from the signal light L, and the autocorrelation function calculation unit 11B multiplies the frequency spectrum by a window function, thereby arbitrarily An autocorrelation function is calculated by cutting out the bandwidth as an effective section, and inverse Fourier transforming the obtained partial spectrum, and a deconvolution operation is performed using the same window function for this autocorrelation function.

[Effect of the sixth embodiment]
As described above, in the present embodiment, the autocorrelation function calculation unit 11B multiplies the frequency spectrum obtained by the spectrum measurement unit 11A by the window function, so that any one of the passbands of the frequency spectrum is obtained. An autocorrelation function is calculated by extracting part or all as a finite interval, inverse Fourier transform is performed on the obtained partial spectrum, and the resulting autocorrelation function is deconvolved with a window function. It is.
As a result, the OSNR error due to distortion caused by using the window function can be reduced, and the OSNR calculation accuracy can be improved. Further, the accuracy can be improved even when a rectangular window is used, and the OSNR can be calculated by a simpler calculation.

[Seventh Embodiment]
Next, an OSNR calculation apparatus 10 according to the seventh embodiment of the present invention will be described with reference to FIGS. FIG. 11 is an explanatory diagram showing changes in OSNR due to the slope of the frequency spectrum. FIG. 12 is an explanatory diagram illustrating an example of inverse Fourier transform using a linear function. FIG. 13 is an explanatory diagram showing an autocorrelation function of a time waveform when the frequency spectrum has a slope.

  In the fifth and sixth embodiments, the distortion compensation of the autocorrelation function in the case of calculating the autocorrelation function by performing inverse Fourier transform on the frequency spectrum has been described. In this embodiment, frequency spectrum inclination correction will be described.

In the present embodiment, the autocorrelation function acquisition unit 11 is provided with a spectrum measurement unit 11A and an autocorrelation function calculation unit 11B as main functional units, as in FIG. 8 described above.
The spectrum measuring unit 11A has a function of measuring the frequency spectrum of the input signal light L.
The autocorrelation function calculation unit 11B corrects the slope of the frequency spectrum by multiplying or dividing the frequency spectrum obtained by the spectrum measurement unit 11A by a linear function having the slope for each slope selected from the search range. And a function of calculating an autocorrelation function for each inclination by performing inverse Fourier transform on each of the corrected frequency spectra.

  The OSNR calculation unit 12 calculates the OSNR for each autocorrelation function for each inclination obtained by the autocorrelation function acquisition unit 11 and calculates the OSNR indicating the extreme value of these OSNRs as the OSNR of the signal light L. It has the function to do.

  When the frequency spectrum used for calculating the autocorrelation function is inclined, as shown in FIG. 11, the OSNR obtained from the autocorrelation function changes. For example, for a frequency spectrum in which the OSNR was −10 dB when the slope sl = 0, when the slope sl = −10 mw / THz was added, the OSNR calculated from the frequency spectrum changed to −16 dB, and the slope It is shown that the OSNR error increases with an increase.

  In general, the frequency spectrum of signal light is inclined due to factors such as frequency dependence characteristics in an amplifier provided in a transmission line. Such a slope of the frequency spectrum can be regarded as equivalent to multiplying the frequency spectrum by a window function of a linear function. That is, the inverse Fourier transform of the frequency spectrum is a convolution of the autocorrelation function of the time waveform of the signal whose frequency spectrum is not inclined and the inverse Fourier transform of the linear function.

When the linear function is subjected to inverse Fourier transformation, as shown in FIG. From this result, it can be seen that the autocorrelation function of the time waveform of the signal spreads. In consideration of this spread, according to the OSNR calculation method described in the first embodiment, as shown in FIG. 13, the autocorrelation value C0 ′ that is the peak value of the triangular waveform and the autocorrelation value of the extracted point C1 ′ changes as in the following equation (4).

In equation (4), a is a coefficient representing a change in signal strength caused by the slope, b is a coefficient representing a change in noise strength caused by the slope, and α ′ is a change in signal strength caused by the slope at the extracted point. , Β is a coefficient representing the change in noise intensity caused by the slope at the extracted point.
At this time, the OSNR is obtained by the following equation (5).

  Assuming that the frequency spectrum tilts around the frequency at which the peak of the frequency spectrum has no tilt, the tilt of the frequency spectrum can be expressed by multiplying the frequency spectrum by (1 + sl · f). In general, the frequency spectrum is symmetric about the peak. Here, sl represents the slope of the linear function, and f represents the relative frequency with the peak of the frequency spectrum being zero.

  By multiplying a symmetric frequency spectrum by a linear function, an asymmetric frequency spectrum is obtained. When the absolute value of the slope of the linear function is the same and the signs are different, the two spectra have a mirror image relationship. These spectra have the same shape when subjected to inverse Fourier transform. For this reason, when the absolute value of the slope of the linear function is the same, the OSNR calculated from the above-described equation (5) is the same value.

Therefore, the OSNR when the slope of the linear function changes is symmetric about the slope 0 of the linear function. Thereby, when the slope of the linear function is 0, the OSNR becomes an extreme value. The characteristics of FIG. 11 described above are calculated in this way.
From this, the slope of the frequency spectrum is changed in the search range, and the OSNR indicating the extreme value among the OSNRs obtained for each slope can be regarded as the OSNR obtained from the frequency spectrum without the slope.

[Operation of the seventh embodiment]
Next, the operation of the OSNR calculation apparatus 10 according to the present embodiment will be described with reference to FIG. FIG. 14 is a flowchart illustrating OSNR calculation processing according to the seventh embodiment.
The OSNR calculation device 10 executes the OSNR calculation process of FIG. 5 in response to a process start instruction from the operator or an input of the signal light L.

First, the spectrum measuring unit 11A of the autocorrelation function acquiring unit 11 measures a frequency spectrum related to the signal light L from the input signal light L (step 200).
Subsequently, the autocorrelation function calculation unit 11B of the autocorrelation function acquisition unit 11 selects a plurality of slopes sl from a preset search range of the slope (step 201), and the frequency spectrum obtained by the spectrum measurement unit 11A. On the other hand, the slope of the frequency spectrum is corrected for each slope sl by multiplying or dividing by a linear function having the slope sl (step 202).

Next, the autocorrelation function calculation unit 11B calculates an autocorrelation function for each inclination sl by performing inverse Fourier transform on the correction spectrum obtained for each inclination sl (step 203).
Thereafter, the OSNR calculation unit 12 calculates the OSNR for each autocorrelation function for each slope sl obtained by the autocorrelation function acquisition unit 11 (step 204), and the OSNR indicating the extreme value among these OSNRs is calculated as the signal light L. Is calculated as the OSNR (step 205).

[Effect of the seventh embodiment]
As described above, in the present embodiment, the autocorrelation function calculation unit 11B of the autocorrelation function acquisition unit 11 performs, for each slope sl selected from the search range, for the frequency spectrum measured by the spectrum measurement unit 11A. The OSNR calculation unit 12 calculates an autocorrelation function for each inclination sl by multiplying or dividing a linear function having an inclination sl to correct the inclination of the frequency spectrum and inverse Fourier transforming the obtained correction spectrum. The OSNR is estimated for each autocorrelation function for each inclination obtained by the autocorrelation function acquisition unit 11, and the OSNR indicating the extreme value of these OSNRs is calculated as the OSNR of the signal light L.

  As a result, even if an inclination occurs in the frequency spectrum of the input signal light L due to factors such as frequency dependence characteristics in the amplifier provided in the transmission path, the frequency is determined based on the frequency spectrum in which the inclination is corrected. It is possible to calculate a more accurate OSNR in which an error due to a spectrum tilt is suppressed.

[Eighth Embodiment]
Next, an OSNR calculation apparatus 10 according to an eighth embodiment of the present invention will be described with reference to FIGS. 15 and 16. FIG. 15 is a block diagram illustrating a configuration of the OSNR calculation apparatus 10 according to the eighth embodiment. FIG. 16 is a block diagram showing the configuration of the autocorrelation function measurement unit.

  In the fourth embodiment, as a specific example of the autocorrelation function acquisition unit 11, a time waveform autocorrelation function is calculated by performing inverse Fourier transform on the frequency spectrum measured from the input signal light L. Explained. In the present embodiment, a case where the autocorrelation function of the time waveform is directly measured from the input signal light L will be described.

  In the present embodiment, the autocorrelation function acquisition unit 11 is provided with an autocorrelation function measurement unit 20 that directly measures the autocorrelation function of the time waveform from the input signal light L. The autocorrelation function measuring unit 20 can be composed of a Michelson, Mach Cender interferometer, or the like. Hereinafter, a case where a Michelson interferometer is used as the autocorrelation function measurement unit 20 will be described as an example.

As shown in FIG. 16, the autocorrelation function measurement unit 20 includes a half mirror 21, a mirror 22, a mirror 23, and a light receiving unit 24 as main functional units.
The half mirror 21 includes a mirror that splits incident light into reflected light and transmitted light at a predetermined ratio, reflects a part of the input signal light L and guides the mirror 22, and the input signal light L. A function of transmitting a part of the light to the mirror 23, a function of transmitting the reflected light from the mirror 22 to the light receiving unit 24, and a function of reflecting the reflected light from the mirror 23 to the light receiving unit 24. .

The mirror 22 is a mirror having a fixed distance from the half mirror 21, and has a function of reflecting the reflected light from the half mirror 21 to the half mirror 21.
The mirror 23 is made of a movable mirror whose distance from the half mirror 21 can be adjusted, and has a function of reflecting the transmitted light from the half mirror 21 to the half mirror 21.
The light receiving unit 24 includes a light receiving sensor that receives incident light and detects the intensity thereof, and detects the intensity of interference that has arrived from the half mirror 21 and is caused by reflected light from the mirror 22 and reflected light from the mirror 22. have.

[Operation of Eighth Embodiment]
Next, the operation of the autocorrelation function measurement unit 20 according to the present embodiment will be described with reference to FIG.
The input signal light L is split into reflected light and transmitted light by the half mirror 21. Of these, the reflected light is reflected by the mirror 22, and then a part of the reflected light passes through the half mirror 21 and enters the light receiving unit 24. On the other hand, the transmitted light from the half mirror 21 is reflected by the mirror 23, and then a part thereof is reflected by the half mirror 21 and enters the light receiving unit 24.

In the light receiving unit 24, the intensity at which the two lights, which are the reflected light from the mirror 22 and the reflected light from the mirror 22, that have arrived from the half mirror 21 interfere with each other, is measured. At this time, the optical path length difference between the two lights can be changed by changing the distance of the mirror 23 with respect to the half mirror 21. Therefore, when the light intensity when this optical path length difference is changed is plotted, an autocorrelation function of a time waveform is obtained. This autocorrelation function is expressed by the following equation (6).

  In equation (6), R is an autocorrelation value, P is the amplitude of the signal, t is time, and τ is a value obtained by dividing the speed of light by the optical path length difference, that is, a value obtained by converting the optical path length difference into time. is there. A similar effect can be obtained with a Mach Cender interferometer, a Sagnac interferometer, or the like.

[Effect of the eighth embodiment]
As described above, in the present embodiment, the autocorrelation function acquisition unit 11 is configured by the autocorrelation function measurement unit 20 that directly measures the autocorrelation function of the time waveform from the input signal light L. There is no need to perform an inverse Fourier transform for calculating the autocorrelation function of the time waveform. For this reason, the calculation processing load in the OSNR calculation apparatus 10 can be greatly reduced.

[Extended embodiment]
The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  DESCRIPTION OF SYMBOLS 10 ... OSNR calculation apparatus, 11 ... Auto correlation function acquisition part, 11A ... Frequency spectrum measurement part, 11B ... Auto correlation function calculation part, 12 ... OSNR calculation part, 20 ... Auto correlation function measurement part, 21 ... Half mirror, 22 ... Mirror, 23 ... Mirror (movable), 24 ... Light receiving part, S ... Signal intensity, N ... Noise intensity, OSNR ... Optical signal to noise ratio.

Claims (8)

An autocorrelation function acquisition step of acquiring an autocorrelation function of a time waveform related to the signal light from the input signal light;
The peak value of the autocorrelation function is a sum of signal intensity and noise intensity, and the signal intensity is estimated based on an intensity value (autocorrelation value) other than the peak value of the autocorrelation function, and the intensity An OSNR calculating step of obtaining the noise intensity by subtracting the signal intensity from a sum, and calculating an OSNR of the signal light from the signal intensity and the noise intensity , and
In the OSNR calculation step, a correlation value at a set time position different from the peak value in the autocorrelation function is obtained in advance from an autocorrelation function of reference light having a known OSNR, and the self-correlation function at the set time position is obtained. An OSNR calculation method , wherein the signal intensity is calculated by dividing by a constant indicating a ratio between a correlation value and a signal intensity of the reference light . An autocorrelation function acquisition step of acquiring an autocorrelation function of a time waveform related to the signal light from the input signal light;
The peak value of the autocorrelation function is a sum of signal intensity and noise intensity, and the signal intensity is estimated based on an intensity value (autocorrelation value) other than the peak value of the autocorrelation function, and the intensity An OSNR calculating step of obtaining the noise intensity by subtracting the signal intensity from a sum and calculating an OSNR of the signal light from the signal intensity and the noise intensity;
With
The OSNR calculation step specifies an approximate line of the autocorrelation function from a plurality of autocorrelation values that are extracted from the autocorrelation function and are different from the peak value and do not sandwich the peak value, The OSNR calculation method, wherein the signal intensity is a value at a time position of the peak value in the approximate line. In the OSNR calculation method according to claim 1 or 2 ,
The autocorrelation function acquisition step includes
A frequency spectrum measuring step for measuring a frequency spectrum of the signal light;
By cutting out a part or all of any one pass band from the frequency spectrum as a finite section, and adding a temporary intensity value outside the finite section, by performing an inverse Fourier transform, An OSNR calculation method comprising: an autocorrelation function calculation step for calculating an autocorrelation function. In the OSNR calculation method according to claim 1 or 2 ,
The autocorrelation function acquisition step includes
A frequency spectrum measuring step for measuring a frequency spectrum of the signal light;
The autocorrelation is performed by performing a deconvolution operation of the window function on the data obtained by inverse Fourier transform after multiplying the frequency spectrum by a window function having any one pass band as a finite section. An OSNR calculation method comprising: an autocorrelation function calculation step for calculating a function. In the OSNR calculation method according to claim 1 or 2 ,
The autocorrelation function acquisition step includes
A frequency spectrum measuring step for measuring a frequency spectrum of the signal light;
For each slope selected from the search range for the frequency spectrum, after correcting the slope of the frequency spectrum by multiplying or dividing by a linear function having the slope, the inverse Fourier transform is performed for each slope. An autocorrelation function calculating step for calculating an autocorrelation function, and
In the OSNR calculation step, the OSNR is calculated for each autocorrelation function for each inclination, and an OSNR indicating an extreme value of the OSNR is used as the OSNR of the signal light. An autocorrelation function acquisition unit that acquires an autocorrelation function of a time waveform related to the signal light from the input signal light;
The peak value of the autocorrelation function is a sum of signal intensity and noise intensity, and the signal intensity is estimated based on an intensity value (autocorrelation value) other than the peak value of the autocorrelation function, and the intensity An OSNR calculation unit that obtains the noise intensity by subtracting the signal intensity from a sum, and calculates an OSNR of the signal light from the signal intensity and the noise intensity ;
The OSNR calculation unit obtains, in advance, a correlation value at a set time position different from the peak value of the autocorrelation function from an autocorrelation function of a reference light having a known OSNR. An OSNR calculation apparatus , wherein the signal intensity is calculated by dividing by a constant indicating a ratio between a correlation value and a signal intensity of the reference light . An autocorrelation function acquisition unit that acquires an autocorrelation function of a time waveform related to the signal light from the input signal light;
The peak value of the autocorrelation function is a sum of signal intensity and noise intensity, and the signal intensity is estimated based on an intensity value (autocorrelation value) other than the peak value of the autocorrelation function, and the intensity An OSNR calculation unit that obtains the noise intensity by subtracting the signal intensity from a sum, and calculates an OSNR of the signal light from the signal intensity and the noise intensity;
With
The OSNR calculation unit identifies an approximate line of the autocorrelation function from a plurality of autocorrelation values that are extracted from the autocorrelation function and that are different from the peak value and that do not sandwich the peak value. An OSNR calculation apparatus, wherein the signal intensity is a value at a time position of the peak value in the approximate line. In the OSNR calculation apparatus according to claim 6 or 7 ,
The autocorrelation function acquisition unit
A frequency spectrum measurement unit for measuring the frequency spectrum of the signal light;
By cutting out a part or all of any one pass band from the frequency spectrum as a finite section, and adding a temporary intensity value outside the finite section, by performing an inverse Fourier transform, An OSNR calculation apparatus comprising: an autocorrelation function calculation unit that calculates an autocorrelation function.

Images