Dispersion measurement device and method

Abstract

This dispersion measurement device 10 includes: a distributed feedback (DFB) laser 11; an electric field absorption (EA) modulator 15 that applies pulse amplitude modulation (PAM) to the laser light supplied from the DFB laser 11, to form an optical signal; a photoelectric (O / E) converter 17 that receives an optical signal propagated through an optical fiber 31; and a signal processing unit 20 that, with respect to the eye pattern of the optical signal received by the O / E converter 17, determines whether the dispersion of the optical fiber 31 is positive or negative on the basis of the displacement of the level of the inner eye so as to make the height of the opening of the inner eye asymmetrical, the dispersion measurement device determining whether the dispersion of the optical fiber is positive or negative.

Description

Dispersion measurement device and method

[0001] The present disclosure relates to dispersion measurement devices and methods.

[0002] In recent years, the demand for increased communication capacity due to the increase in communication traffic has necessitated high-speed optical signal transmission. Pulse amplitude modulation (PAM), which modulates the intensity of light, has achieved high-speed signal transmission exceeding 50 Gbaud using an electro-absorption (EA) modulator. However, light intensity modulation using an EA modulator produces fluctuations in the optical frequency called chirp. Optical fiber, which serves as a transmission path, exhibits dispersion, meaning that the propagation speed varies depending on the frequency of the propagating light. This causes changes in the pulse waveform when intensity-modulated optical pulses propagate through the optical fiber. This can lead to interference between successive pulses, resulting in errors in determining the transmitted signal.

[0003] F. Dorgeuille and F. Devaux, “On the transmission performances and the chirp parameter of a multiple-quantum-well electroabsorption modulator,” in IEEE Journal of Quantum Electronics, vol. 30, no. 11, pp. 2565-2572, Nov. 1994, doi: 10.1109 / 3.333708.

[0004] In direct modulation / direct detection (IMDD) transmission of high-speed optical signals exceeding 25 Gbaud, the effect of optical fiber dispersion is significant, so signal transmission at wavelengths known as the O-band, which includes the zero-dispersion wavelength of commonly used single-mode optical fibers, is the mainstream. The zero-dispersion wavelength is generally considered to be around 1310 nm, but it fluctuates depending on the structure of the waveguide, stress on the fiber, environmental temperature, and other factors. The wavelengths shorter than the zero-dispersion wavelength are called the negative dispersion region, and the shorter the wavelength, the lower the group velocity. On the other hand, the wavelengths longer than the zero-dispersion wavelength are called the positive dispersion region, and the longer the wavelength, the lower the group velocity.

[0005] In an EA modulator, it is possible to change the chirp by changing the bias superimposed on the modulated signal. Therefore, if the dispersion region of the transmission path is known, it is possible to improve the transmission characteristics by controlling the applied bias. Non-Patent Document 1 describes that the chirp characteristic with respect to the bias of an EA modulator is positive in the region where the bias voltage is low and negative in the region where the bias voltage is high.

[0006] However, the zero-dispersion wavelength of the installed optical fiber was unknown, and it was also unknown whether the dispersion of the optical fiber at the wavelength of the signal light was positive or negative. Therefore, it was not possible to determine whether the bias of the EA modulator should be set to a positive or negative condition.

[0007] The present disclosure has been proposed in view of the above-described circumstances, and aims to provide a dispersion measurement device and method that can determine whether the dispersion of an optical fiber is positive or negative.

[0008] In order to solve the above-mentioned problems, the dispersion measurement device disclosed herein measures the dispersion of an optical fiber and includes a laser light source, an electro-absorption modulator that performs pulse amplitude modulation on laser light supplied from the laser light source to generate an optical signal, an opto-electrical converter that receives the optical signal that has propagated through the optical fiber, and a signal processing unit that determines whether the dispersion of the optical fiber is positive or negative based on a displacement of the level of the inner eye in the eye pattern of the optical signal received by the opto-electrical converter that makes the height of the inner eye opening asymmetric.

[0009] The dispersion measurement method disclosed herein measures the dispersion of an optical fiber, and includes pulse amplitude modulation of laser light using an electroabsorption modulator to generate an optical signal, photoelectrically converting the optical signal that has propagated through the optical fiber, and determining whether the dispersion of the optical fiber is positive or negative based on a displacement of the inner eye level in the eye pattern of the photoelectrically converted optical signal that makes the height of the inner eye opening asymmetric.

[0010] In addition, the dispersion measurement device of the present disclosure may be for measuring the dispersion of an optical fiber and may include a laser light source, an electro-absorption modulator that generates an optical signal by pulse amplitude modulating the laser light supplied from the laser light source, an opto-electrical converter that receives the optical signal that has propagated through the optical fiber, and a signal processing unit that determines whether the dispersion of the optical fiber is positive or negative based on the optical intensity of multiple valleys in a histogram of the optical intensity of the optical signal received by the opto-electrical converter.

[0011] The dispersion measurement method disclosed herein is for measuring the dispersion of an optical fiber, and may include pulse amplitude modulating laser light using an electroabsorption modulator to generate an optical signal, photoelectrically converting the optical signal propagated through the optical fiber, and determining whether the dispersion of the optical fiber is positive or negative based on the optical intensity of multiple valleys in a histogram of the optical intensity of the optical signal received by the photoelectric converter.

[0012] According to the present disclosure, it is possible to determine whether the dispersion of an optical fiber is positive or negative.

[0013] 1 is a block diagram showing a schematic configuration of a dispersion measurement device according to a first embodiment; FIG. 2 is a graph showing the relationship between the bias applied to an EA modulator and the intensity of transmitted light; FIG. 3 is a graph showing the relationship between the bias applied to the EA modulator and chirp; FIG. 4 is an eye pattern of the light intensity before and after propagation through an optical fiber; FIG. 5 is a diagram explaining a series of processing steps in a signal processing unit; FIG. 6 is a diagram explaining the process of creating a discretized histogram from the time waveform of an optical signal; FIG. 7 is a diagram explaining peak detection and normalization of an optical signal; FIG. 8 is a graph showing the relationship between the normalized intensity of the inner eye level and dispersion; FIG. 9 is a block diagram showing a schematic configuration of a dispersion measurement device according to a second embodiment; FIG. 10 is a graph showing the relationship between the bias applied to an EA modulator and the intensity of transmitted light; FIG. 11 is a histogram of the intensity of transmitted light; FIG. 12 is a graph showing the relationship between the bias applied to the EA modulator and chirp; FIG. 13 is an eye pattern of the light intensity before and after propagation through an optical fiber; FIG. 14 is a diagram showing the first half of the processing in a signal processing unit; FIG. 15 is a histogram of discretized intensity; FIG. 16 is a diagram explaining the second half of the processing in a signal processing unit; FIG. 17 is a diagram explaining normalization of a histogram; FIG. 18 is a graph showing the relationship between the rate of change of the intensity of a second valley V12′ and dispersion. 10 is a diagram showing the latter half of the process in the signal processing unit of the modified example.FIG. 11 is a graph showing the result of fitting a sum function of Causs functions to a normalized histogram.

[0014] First Embodiment The dispersion measurement device and method of the present disclosure will be described in detail below with reference to the drawings. In the first embodiment, a pulse amplitude modulated (PAM) optical signal is generated using an electroabsorption (EA) modulator, and whether the dispersion of the optical fiber under test (FUT) is positive or negative is determined based on the change in the waveform of the optical signal after propagation through the optical fiber under test (FUT). In the first embodiment, the PAM is assumed to be four-level PAM4 with a signal rate of 50 Gbaud, but other PAM modulations such as PAM8 and other signal rates can also be applied.

[0015] 1 is a block diagram showing a schematic configuration of a dispersion measurement apparatus 10 according to a first embodiment. The dispersion measurement apparatus 10 includes a distributed feedback (DFB) laser 11 serving as a laser light source, a PAM signal generator 12 that generates a PAM signal, a DC bias voltage generator 13 that generates a direct current (DC) bias voltage, and an AC / DC signal superimposer 14 that superimposes the PAM signal, which is an alternating current (AC) signal supplied from the PAM signal generator 12, on the DC bias voltage supplied from the DC bias voltage generator 13.

[0016] The dispersion measurement device 10 also includes an EA modulator 15 that modulates laser light supplied from the DFB laser 11 with an electrical signal supplied from the AC / DC signal superimposer 14 to generate an optical signal, and an optical amplifier 16 that amplifies the optical signal output from the EA modulator 15. The optical signal output from the optical amplifier 16 is guided to an optical fiber under test (FUT) 31 via an optical path 30. The EA modulator 15 has a characteristic in which its transmission loss changes depending on the applied bias voltage, and this characteristic is utilized to achieve optical modulation. The optical fiber 31 is a single-mode (SM) fiber with inherent dispersion. The optical amplifier 16 is provided to improve the signal-to-noise ratio (SNF) of the optical signal propagating through the optical fiber 31 and is composed of a semiconductor optical amplifier (SOA).

[0017] 2 is a graph showing the relationship between the bias voltage applied to the EA modulator 15 and the intensity of transmitted light. The EA modulator 15 generates an optical signal by modulating the intensity of the laser light supplied from the DFB laser 11 in accordance with the bias voltage. In the first embodiment, the 0th level voltage V Ｌ０ , the first level voltage V Ｌ１ , the second level voltage V Ｌ２ and the third level voltage V Ｌ３By alternately applying these four bias voltages to the EA modulator 15, a PAM4 optical signal is generated in which the optical pulse is modulated at four levels of intensity, from the lowest level L0 to the third level L3. The PAM4 optical signal has peak intensities at each of the four levels, from the lowest level L0 to the third level L3.

[0018] The 0th level voltage V applied to the EA modulator 15 Ｌ０ to the third level voltage V Ｌ３ The four bias voltage levels are set so that the four levels of transmitted light intensity from the 0th level L0 to the 3rd level L3 increase uniformly in a stepped manner. In other words, the transmitted light intensity is set so that the interval between the 0th level L0 and the first level L1, the interval between the first level L1 and the second level L2, and the interval between the second level L2 and the third level L3 are uniform.

[0019] 2 also shows the eye pattern of the PAM4 signal generated by the EA modulator 15. This eye pattern includes three inner eyes: a lower inner eye E1, a middle inner eye E2, and an upper inner eye E3, as well as an outer eye E4. As described above, the four levels of transmitted light intensity, from the 0th level L0 to the 3rd level L3, are set to be equally spaced, so the opening heights of the lower inner eye E1, the middle inner eye E2, and the upper inner eye E3 are uniform. In this specification, with reference to this eye pattern, the light intensities of the 0th level L0 and the 3rd level L3 will be referred to as the outer eye levels, and the light intensities of the 1st level L1 and the 2nd level L2 will be referred to as the inner eye levels.

[0020] FIG. 3 is a graph showing the relationship between the bias voltage applied to the EA modulator 15 and chirp. The EA modulator 15 typically has an absorption layer with a multiple quantum well (MQW) structure, and modulates light intensity by utilizing the shift in the absorption edge due to the quantum-confined Stark effect. During this process, a refractive index change occurs as expressed by the Kramers-Kronig (K-K) relationship, modulating the phase of the light and resulting in frequency chirp. As seen in Non-Patent Document 1, the refractive index change with respect to the bias voltage is parabolic, with the refractive index increasing with increasing bias in the low bias voltage region and decreasing with increasing bias in the high bias voltage region. Considering this as frequency chirp, the low bias region exhibits a positive chirp characteristic, while the high bias region exhibits a negative chirp characteristic, as shown in FIG. 3.

[0021] The zeroth level voltage V is a four-value bias voltage corresponding to the PAM4 optical signal. Ｌ０ to the third level voltage V Ｌ３ The chirp parameters corresponding to the four bias values ​​are the 0th level chirp parameters α Ｌ０ to the third-level chirp parameter α Ｌ３ The first-level chirp parameter α Ｌ１ to the third-level chirp parameter α Ｌ３ In a positive chirp such as Ｌ０ In a negative chirp such as this, the rising edge of the pulse is red-shifted and the falling edge is blue-shifted.

[0022] 1 , the dispersion measurement device 10 further includes an O / E converter 17 that performs opto-electric (O / E) conversion of an optical signal to an electrical signal, an A / D converter 18 that performs analog-to-digital (A / D) conversion of the electrical signal supplied from the O / E converter 17, a memory 19 that stores data supplied from the A / D converter 18, and a signal processing unit 20 that receives the data stored in the memory 19 and determines whether the dispersion of the optical fiber 31 is positive or negative. The O / E converter 17 receives the optical signal that has propagated through the optical fiber 31 via an optical path 30. The optical signal that enters the O / E converter 17 is converted into an electrical signal, sampled by the A / D converter 18, and converted into binary data that is stored in the memory 19. The signal processing unit 20 receives the binary data from the memory 19.

[0023] FIG. 4 shows the eye patterns of an optical signal before and after propagation through the optical fiber 31. It is known that chirp affects the pulse waveform when an optical signal propagates through a dispersive medium such as the optical fiber 31. Because the group velocity differs depending on the wavelength of light in a dispersive medium, the waveform of a chirped pulse is distorted after propagation. For this reason, the eye pattern waveform of the PAM4 optical signal generated by the EA modulator 15 changes before and after propagation through the optical fiber 31, as shown in FIG. 4. Note that in FIG. 4, the outer levels, 0th level L0 and 3rd level L3, of each eye pattern are aligned.

[0024] The center of Fig. 4 shows the eye pattern of the PAM4 optical signal before it propagates through the optical fiber 31. This eye pattern is due to the PAM4 optical signal generated by the EA modulator 15 and before it enters the optical fiber 31 from the optical amplifier 16 via the optical path 30. Because the optical signal before it propagates through the optical fiber 31 is not affected by dispersion due to the optical fiber 31, it has the same eye pattern as the PAM4 optical signal generated by the EA modulator 15 shown in Fig. 2, and the opening heights of the lower inner eye E0, middle inner eye E1, and upper inner eye E3 are uniform.

[0025] The left side of Figure 4 shows the eye pattern of an optical signal after propagation through an optical fiber 31 with negative dispersion. When the optical fiber 31 has negative dispersion, a positively chirped pulse has a low group velocity at the rising edge where a blue shift occurs and a high group velocity at the falling edge where a red shift occurs. Propagation is delayed at the rising edge and accelerated at the falling edge, narrowing the pulse width on the time axis as if compressed from both sides, compressing the pulse and increasing its peak intensity. On the other hand, a negatively chirped pulse, on the other hand, propagates faster at the rising edge where a red shift occurs and slower at the falling edge where a blue shift occurs, broadening the pulse and decreasing its peak intensity. Therefore, when the optical fiber 31 has negative dispersion, pulse compression occurs in PAM4 symbols with high optical intensity, increasing the peak intensity of the pulse and widening the eye pattern opening, resulting in an asymmetric eye pattern in which the height of the opening of the upper inner eye E3 is greater than the height of the opening of the lower inner eye E1.

[0026] The right side of Figure 4 shows the eye pattern of an optical signal after propagation through an optical fiber 31 with positive dispersion. When the optical fiber 31 has positive dispersion, a positively chirped pulse has a high group velocity at the rising edge where a blue shift occurs and a low group velocity at the falling edge where a red shift occurs. Because propagation is fast at the rising edge and slow at the falling edge, the pulse width widens in the time domain and the peak intensity of the pulse decreases. On the other hand, a negatively chirped pulse, on the other hand, has a slower propagation at the rising edge where a red shift occurs and a faster propagation at the falling edge where a blue shift occurs. Therefore, the pulse width narrows on the time axis as if the pulse were compressed from both sides, and the pulse is compressed and its peak intensity increases. Therefore, when the optical fiber 31 has positive dispersion, the optical fiber 31 exhibits the opposite tendency to when it has negative dispersion. For PAM4 symbols with high optical intensity, the pulse widens, the peak intensity of the pulse decreases, and the eye pattern opening narrows, resulting in an asymmetric eye pattern in which the height of the opening of the lower inner eye E1 is greater than the height of the opening of the upper inner eye E3.

[0027] 5 is a diagram illustrating a series of processing steps in the signal processing unit 20. The signal processing unit 20 performs the following series of steps to determine whether the dispersion of the optical fiber 31 is positive or negative based on the data of the optical signal provided from the memory 19.

[0028] In the first step S1, the signal processing unit 20 receives optical signal data in a binary data array, which is binary data, from the memory 19 and converts it into a decimal data array in decimal notation. When the number of quantization bits of the A / D converter 18 is 8 bits, the decimal data array takes integer values ​​in the range of 0 to 255.

[0029] In step S2, the signal processing unit 20 creates a histogram of the optical signal discretized into integer values. Regarding step S2, referring to FIG. 6 , the time waveform of the optical signal is discretized by sampling at a predetermined sampling interval T. The histogram is generated by counting the discretized integer values. It can be observed that the histogram has four peaks corresponding to the zeroth level L0 to the third level L3 of the optical intensity of the optical signal. The counts in the histogram may be expressed logarithmically, and the logarithm may have a base of 2.

[0030] In step S3, the signal processing unit 20 detects peaks from the histogram. Referring to FIG. 7, the left side shows a histogram obtained by discretizing the time waveform of the optical signal, as shown in FIG. 6. The distribution of counts centered on the peak generally takes a normal distribution based on a Gaussian function, since it is due to thermal noise and shot noise. Therefore, the distribution function of the entire histogram can be assumed to be the sum of four Gaussian functions. By detecting peaks by fitting the Gaussian functions, the discretized intensities of the four peaks, from the 0th level L0 to the 3rd level L3, can be obtained.

[0031] In step S4, the signal processing unit 20 normalizes the optical intensity based on the intensities of the zeroth level L0 and the third level L3 of the outer eye level. In the first embodiment, normalization is performed so that the intensity of the zeroth level L0 is −3 and the intensity of the third level L3 is +3. Therefore, the intermediate value between the zeroth level L0 and the third level L3 is 0. When the heights of the inner eye openings in the eye pattern are uniform, as in the PAM4 optical signal before propagation through the optical fiber 31 shown in the center of FIG. 4 , that is, when the heights of the openings of the lower inner eye E1, middle inner eye E2, and upper inner eye E2 are uniform, the first level L1, which is the inner eye level, is −1, and the second level L2 is 1.

[0032] In step S5, the signal processing unit 20 determines whether the dispersion of the optical fiber 31 is positive or negative. As described with reference to Fig. 4, when the dispersion of the optical fiber 31 is zero, the eye pattern of the optical signal does not change before and after propagation through the optical fiber 31, and the opening heights of the lower inner eye E1, middle inner eye E2, and upper inner eye E3 are uniform, similar to the eye pattern before propagation through the optical fiber 31 shown in the center of Fig. 4. The normalized intensities of the first level L1 and second level L2, which are the inner eye levels, also do not change before and after propagation through the optical fiber 31, with the first level L1 being -1 and the second level L2 being +1.

[0033] When the optical fiber 31 has negative dispersion, the eye pattern of the optical signal after propagation through the optical fiber 31 becomes asymmetric, with the height of the opening of the upper inner eye E3 being greater than the height of the opening of the lower inner eye E1, as shown on the left side of Fig. 4. Since the normalized intensity of the optical signal is normalized so that the zeroth level L1 and the third level L3 are -3 and +3, the first level L1 and the second level L2 are lower than the values ​​of -1 and +1 of the first level L1 and the second level L2, respectively, of the eye pattern before propagation through the optical fiber 31 shown in the center of Fig. 4.

[0034] When the optical fiber 31 has positive dispersion, the eye pattern of the optical signal after propagation through the optical fiber 31 becomes asymmetric, with the height of the opening of the lower inner eye E1 being greater than the height of the opening of the upper inner eye E3, as shown on the right side of Fig. 4. The normalized intensity of the optical signal is normalized so that the zeroth level L1 and the third level L3 are -3 and +3, and therefore the first level L1 and the second level L2 are higher than the values ​​of -1 and +1 of the first level L1 and the second level L2, respectively, of the eye pattern before propagation through the optical fiber 31 shown in the center of Fig. 4.

[0035] It is possible to determine whether the dispersion of the optical fiber 31 is positive or negative based on the eye pattern of the optical signal after propagation through such an optical fiber 31. As shown in the center of Figure 4, similar to the eye pattern before propagation, when the opening heights of the lower inner eye E1, middle inner eye E2, and upper inner eye E3 are uniform, it is possible to estimate that the dispersion of the optical fiber 31 is zero. Regarding the normalized intensities of the first level L1 and second level L2 of the inner eye levels, it is possible to estimate that the dispersion of the optical fiber 31 is zero when the first level L1 is -1 and the second level L2 is +1.

[0036] 4, when the height of the opening of the upper inner eye E3 is greater than the height of the opening of the lower inner eye E1, resulting in an asymmetrical relationship, it can be determined that the dispersion of the optical fiber 31 is negative. Regarding the normalized intensities of the first level L1 and the second level L2 of the inner eye levels, when the first level L1 is lower than −1 and the second level L2 is lower than +1, it can be estimated that the dispersion of the optical fiber is negative.

[0037] 4, when the height of the opening of the lower inner eye E1 is greater than the height of the opening of the upper inner eye E3 and is asymmetric, it can be determined that the dispersion of the optical fiber 31 is positive. Regarding the normalized intensities of the first level L1 and the second level L2 of the inner eye levels, when the first level L1 is higher than −1 and the second level L2 is higher than +1, it can be estimated that the dispersion of the optical fiber 31 is positive.

[0038] 8 is a graph showing the relationship between dispersion and normalized intensity of the first and second inner eye levels L1 and L2. This relationship between the inner eye level, normalized intensity, and dispersion depends on the characteristics of the EA modulator 15, and can therefore be created by measuring the inner eye level of known optical fibers 31 with different dispersions. This relationship is created by the signal processing unit 20 and stored in the memory 19.

[0039] Dispersion measurement apparatus 10 can measure the dispersion of optical fiber 31 by referring to the relationship between the normalized intensity of the inner eye level and the dispersion stored in memory 19. Signal processing unit 20 evaluates the dispersion of the normalized intensity of the optical signal obtained by the normalization process in step S4 of Fig. 5 by referring to the relationship between the normalized intensity and the dispersion stored in memory 19. With reference to the relationship between the normalized intensity and the dispersion as shown in Fig. 8, for example, when the normalized intensity of the first level L1 and the second level L2 of the inner eye level of the optical signal propagated through optical fiber 31 changes by -20% compared to the optical signal propagated through optical fiber 31 with zero dispersion, the amount of dispersion of optical fiber 31 can be estimated to be -50 ps / nm.

[0040] As described above, according to the first embodiment, it is possible to determine whether the dispersion of the optical fiber is positive or negative based on a shift in the level of the inner eye that causes the height of the inner eye opening to become asymmetric with respect to the eye pattern of an optical signal transmitted through the optical fiber 31. Therefore, even for an installed optical fiber whose zero-dispersion wavelength is unknown, it is possible to optimize performance by determining whether the dispersion is positive or negative.

[0041] Furthermore, according to the first embodiment, it is possible to estimate not only whether the dispersion of the optical fiber 31 is positive or negative but also the amount of dispersion by further using the known characteristics of the EA modulator 15. Since the dispersion of the optical fiber 31 can be determined by directly analyzing the PAM signal used in actual signal transmission, this method can be applied to optical transceivers.

[0042] The signal processing unit 20 and memory 19 of the dispersion measurement device 10 described above can be, for example, a general-purpose computer system. The computer system includes a CPU (Central Processing Unit, or processor), memory, storage (HDD: Hard Disk Drive, SSD: Solid State Drive), a communication device, an input device, an output device, and the like. The memory and storage are storage devices. In this computer system, the CPU executes a predetermined program loaded into the memory, thereby realizing each function of the signal processing unit 20. The program for the signal processing unit 20 can be stored in a computer-readable recording medium such as an HDD, SSD, USB (Universal Serial Bus) memory, CD (Compact Disc), or DVD (Digital Versatile Disc), or can be distributed via a network. The computer-readable recording medium is, for example, a non-transitory recording medium.

[0043] It should be noted that the present disclosure is not limited to the first embodiment, and various modifications are possible within the scope of the gist thereof.

[0044] Second Embodiment Next, a second embodiment of the dispersion measurement apparatus and method of the present disclosure will be described. In the second embodiment, a pulse amplitude modulated (PAM) optical signal is generated using an electroabsorption (EA) modulator, and whether the dispersion of the optical fiber is positive or negative is determined from the waveform change of the optical signal after propagation through the optical fiber under test (FUT). In the second embodiment, the PAM is assumed to be four-level PAM4 with a signal rate of 50 Gbaud, but this is not limiting and other PAM modulations such as PAM8 and other signal rates can also be applied.

[0045] 9 is a block diagram showing a schematic configuration of a dispersion measurement apparatus 110 according to the second embodiment. The dispersion measurement apparatus 110 includes a distributed feedback (DFB) laser 111 serving as a laser light source, a PAM signal generator 112 that generates a PAM signal, a DC bias voltage generator 113 that generates a direct current (DC) bias voltage, and an AC / DC signal superimposer 114 that superimposes the PAM signal, which is an alternating current (AC) signal supplied from the PAM signal generator 112, on the DC bias voltage supplied from the DC bias voltage generator 113.

[0046] The dispersion measurement device 110 also includes an EA modulator 115 that modulates laser light supplied from the DFB laser 111 with an electrical signal supplied from an AC / DC signal superimposer 114 to generate an optical signal, and an optical amplifier 116 that amplifies the optical signal output from the EA modulator 115. The optical signal output from the optical amplifier 116 is guided to an optical fiber under test (FUT) 131 via an optical path 130. The EA modulator 115 has a characteristic in which its transmission loss changes depending on the applied bias voltage, and this characteristic is utilized to achieve optical modulation. The optical fiber 131 is a single-mode (SM) fiber with inherent dispersion. The optical amplifier 116 is provided to improve the signal-to-noise ratio (SNF) of the optical signal propagating through the optical fiber 131 and is configured as a semiconductor optical amplifier (SOA).

[0047] 10 is a graph showing the relationship between the bias voltage applied to the EA modulator 115 and the intensity of transmitted light. The EA modulator 115 generates an optical signal by modulating the optical intensity of the laser light supplied from the DFB laser 111 in accordance with the bias voltage. In the second embodiment, the 0th level voltage V Ｌ０ , the first level voltage V Ｌ１ , the second level voltage V Ｌ２ and the third level voltage V Ｌ３By alternately applying these four bias voltage levels to the EA modulator 115, a PAM4 optical signal is generated in which the optical pulse is modulated at four levels of intensity, from the lowest level L0 to the third level L3. The PAM4 optical signal has peak intensities at each of the four levels, from the lowest level L0 to the third level L3.

[0048] The 0th level voltage V applied to the EA modulator 115 Ｌ０ to the third level voltage V Ｌ３ The four bias voltage levels are set so that the four levels of transmitted light intensity from the 0th level L0 to the 3rd level L3 increase uniformly in a stepped manner. In other words, the transmitted light intensity is set so that the interval between the 0th level L0 and the first level L1, the interval between the first level L1 and the second level L2, and the interval between the second level L2 and the third level L3 are uniform.

[0049] 10 also shows the eye pattern of the PAM4 signal generated by the EA modulator 115. The eye pattern is obtained by continuously acquiring a signal waveform over a certain period of time, dividing the acquired signal into 1-bit period units called unit intervals (UI), and overwriting each divided UI within a range of 1 UI.

[0050] Here, the eye pattern formed by the zeroth level L0 and the third level L3 is called the outer eye E4, and the eye pattern formed by the first level L1 and the second level L2 is called the inner eye. There are two other inner eyes: a first inner eye E1 formed by the first level L0 and the first level L1, a second inner eye E2 formed by the first level L1 and the second level L2, and a third inner eye E3 formed by the second level L2 and the third level L3. The four levels of transmitted light intensity from the zeroth level L0 to the third level L3 are set to be equally spaced, so the opening heights of the lower inner eye E1, middle inner eye E2, and upper inner eye E3 are uniform.

[0051] Fig. 11 is a histogram of the intensity of the PAM4 signal generated by the EA modulator 115. Fig. 11 also shows the eye pattern of the PAM4 signal, as in Fig. 10. The histogram in Fig. 11 can be created by sampling at the center of the unit interval of the eye pattern and counting the intensity of each sample.

[0052] FIG. 12 is a graph showing the relationship between the bias applied to the EA modulator 115 and chirp. The EA modulator 115 typically has an absorption layer with a multiple quantum well (MQW) structure, and modulates light intensity by utilizing the shift in the absorption edge due to the quantum-confined Stark effect. This produces a refractive index change expressed by the Kramers-Kronig (K-K) relationship, modulating the phase of the light and resulting in frequency chirp. As seen in Non-Patent Document 1, the refractive index change with respect to the bias voltage is parabolic, with the refractive index increasing with increasing bias in the low bias voltage region and decreasing with increasing bias in the high bias voltage region. Considering this as frequency chirp, the low bias region exhibits a positive chirp characteristic, while the high bias region exhibits a negative chirp characteristic, as shown in FIG. 12.

[0053] The zeroth level voltage V is a four-value bias voltage corresponding to the PAM4 optical signal. Ｌ０ to the third level voltage V Ｌ３ The chirp parameters corresponding to the four bias values ​​are the chirp parameters α Ｌ０ From the chirp parameter α of the third level L3 Ｌ３ The chirp parameter α of the first level L1 is different from Ｌ１ From the chirp parameter α of the third level L3 Ｌ３ In a positive chirp such as Ｌ０ In a negative chirp such as this, the rising edge of the pulse is red-shifted and the falling edge is blue-shifted.

[0054] 9 , the dispersion measurement device 110 further includes an O / E converter 117 that performs opto-electric (O / E) conversion on the optical signal to generate an electrical signal, an A / D converter 118 that performs analog-to-digital (A / D) conversion on the electrical signal supplied from the O / E converter 117, a memory 119 that stores data supplied from the A / D converter 118, and a signal processing unit 120 that receives the data stored in the memory 119 and determines whether the dispersion of the optical fiber 131 is positive or negative. The O / E converter 117 receives the optical signal that has propagated through the optical fiber 131 via an optical path 130. The optical signal that enters the O / E converter 117 is converted into an electrical signal, sampled by the A / D converter 118, and converted into binary data that is stored in the memory 119. The signal processing unit 120 receives the binary data from the memory 119.

[0055] Fig. 13 shows the eye patterns of an optical signal before and after propagation through the optical fiber 131. It is known that chirp affects the pulse waveform when an optical signal propagates through a dispersive medium such as the optical fiber 131. Because the group velocity differs depending on the wavelength of light in a dispersive medium, the waveform of a chirped pulse is distorted after propagation. For this reason, the eye pattern waveform of the PAM4 optical signal generated by the EA modulator 115 changes before and after propagation through the optical fiber 131, as shown in Fig. 13. Note that in Fig. 13, the outer levels, 0th level L0 and 3rd level L3, of each eye pattern are aligned.

[0056] The center of Fig. 13 shows the eye pattern of the PAM4 optical signal before it propagates through the optical fiber 131. This eye pattern is due to the PAM4 optical signal generated by the EA modulator 115 and before it enters the optical fiber 131 from the optical amplifier 116 via the optical path 130. The optical signal before it propagates through the optical fiber 131 is not affected by dispersion due to the optical fiber 131, and therefore has the same eye pattern as the PAM4 optical signal generated by the EA modulator 15 shown in Figs. 10 and 11, and the opening heights of the lower inner eye E0, middle inner eye E1, and upper inner eye E3 are uniform.

[0057] The left side of Figure 13 shows the eye pattern of an optical signal after propagation through an optical fiber 131 with negative dispersion. When the optical fiber 131 has negative dispersion, a positively chirped pulse has a low group velocity at the rising edge where a blue shift occurs and a high group velocity at the falling edge where a red shift occurs. Propagation is delayed at the rising edge and accelerated at the falling edge, narrowing the pulse width on the time axis as if compressed from both sides. The pulse is compressed and its peak intensity increases. On the other hand, a negatively chirped pulse, on the other hand, propagates faster at the rising edge where a red shift occurs and slower at the falling edge where a blue shift occurs, broadening the pulse and reducing its peak intensity. Therefore, when the optical fiber 131 has negative dispersion, pulse compression occurs in symbols with high PAM4 optical intensity, increasing the peak intensity of the pulse and widening the eye pattern opening. This results in an asymmetric eye pattern in which the height of the opening of the upper inner eye E3 is greater than the height of the opening of the lower inner eye E1.

[0058] The right side of Figure 13 shows the eye pattern of an optical signal after propagation through an optical fiber 131 with positive dispersion. When the optical fiber 131 has positive dispersion, a positively chirped pulse has a high group velocity at the rising edge where a blue shift occurs and a low group velocity at the falling edge where a red shift occurs. Because propagation is fast at the rising edge and slow at the falling edge, the pulse width widens in the time domain and the peak intensity of the pulse decreases. On the other hand, a negatively chirped pulse, on the other hand, has a slower propagation at the rising edge where a red shift occurs and a faster propagation at the falling edge where a blue shift occurs. Therefore, the pulse width narrows on the time axis as if the pulse were compressed from both sides, resulting in a higher peak intensity. Therefore, when the optical fiber 131 has positive dispersion, the optical fiber 131 exhibits the opposite tendency to when it has negative dispersion. For PAM4 symbols with high optical intensity, the pulse widens, the peak intensity of the pulse decreases, and the eye pattern opening narrows, resulting in an asymmetric eye pattern in which the height of the opening of the lower inner eye E1 is greater than the height of the opening of the upper inner eye E3.

[0059] 14 is a diagram illustrating the first half of the process in the signal processing unit 120. The signal processing unit 120 performs the following series of processes to determine whether the dispersion of the optical fiber 131 is positive or negative based on the optical signal data provided from the memory 119. Here, as an example, the dispersion of the optical fiber under test (FUT) 131 is assumed to be 0. In this case, the four signal levels of PAM4 are equally spaced, and the openings of the first inner eye E1, second inner eye E2, and third inner eye E1 are uniform, as in the eye pattern before propagation through the optical fiber 131 shown in the center of FIG.

[0060] In the first step S11, the signal processing unit 120 receives optical signal data in a binary data array, which is binary data, from the memory 119 and converts it into a decimal data array in decimal notation. When the number of quantization bits of the A / D converter 118 is 8 bits, the decimal data array takes integer values ​​in the range of 0 to 255.

[0061] In step S12, the number of samples of each integer value in the data string is counted sequentially. The number of samples is counted sequentially, starting with integers 0, 1, 2, ..., 255, to create a count array with 256 elements. This 2 x 256 matrix represents the histogram information. The histogram shown in Figure 15 is presented as a bar graph in which the horizontal axis represents the discrete intensity of integer values ​​corresponding to the intensity, and the vertical axis represents the count in base 2 logarithm. It can be seen that the histogram has four peaks corresponding to the optical intensity levels L0 to L3 of the optical signal. Note that in the histogram, the count may be displayed logarithmically, with the logarithm base 2.

[0062] 16 is a diagram illustrating the latter half of the process in the signal processing unit 120. In step S13, the signal processing unit 120 performs valley detection on the matrix of histogram information, and detects the intensity of a minimum from the valleys of three count numbers between four count peaks corresponding to each signal level of PAM4.

[0063] 17, in the histogram shown on the left side of FIG. 15, the peak intensities are 45, 101, 150, and 185 for the 0th peak L0, the 1st peak L1, the 2nd peak L2, and the 3rd peak L3, respectively. The valley intensities are 68, 115, and 162 for the 1st valley V01, the 2nd valley V12, and the 3rd valley V23, respectively. The 1st valley V01, the 2nd valley V12, and the 3rd valley V23 are formed between adjacent peaks in the 0th peak L0, the 1st peak L1, the 2nd peak L2, and the 3rd peak L2, respectively. The peak spacing and the valley spacing are equal.

[0064] In step S14, the intensities of the first valley V01 and the third valley V23 are normalized to -1 and +1, respectively. After normalization, the intensities of the first valley V01', the second valley V12', and the third valley V23' become -1, 0, and +1, respectively. As shown in the normalized intensity histogram on the right side of Figure 17, the intensity of the second valley V12, which was 115 before normalization, becomes 0 in the normalized second valley V12'.

[0065] 18 shows the results of calculating the rate of change of the normalized intensity of the second valley V12' when the optical fiber 131 has dispersion. As described above, the dispersion of the optical fiber 131 is assumed to be zero, but optical fiber 131 generally has dispersion. In the second embodiment, the dispersion is determined from the change in the normalized intensity of the second valley V12'. This rate of change is calculated by setting 100% to the normalized intensity of the third valley V23' and -100% to the position of the normalized intensity of the first valley V12'.

[0066] As is clear from Figure 18, in the case of positive dispersion, the normalized intensity of the second valley V12' shifts to the positive side, while in the case of negative dispersion, it shifts to the negative side. This makes it possible to determine the sign of the dispersion of the optical fiber 131 based on whether it shifts to a positive or negative side. The change with respect to dispersion is linear, and the slope depends on the chirp characteristics of the transmitter. Therefore, for a measurement system in which the chirp characteristics are known, it is possible to measure the chirp amount of the optical fiber 131 from the amount of change by previously acquiring the correspondence relationship shown in Figure 18.

[0067] In dispersion measurement device 110, the relationship between dispersion and the normalized intensity of V12' at the second valley depends on the characteristics of EA modulator 115. For this reason, it can be created by measuring the level of the normalized intensity of V12' at the second valley for known optical fibers 131 with different dispersions. This relationship is created by signal processing unit 120 and stored in memory 119.

[0068] Dispersion measurement device 110 can measure the dispersion of optical fiber 131 by referring to the relationship between the dispersion and the normalized intensity of the normalized second valley V12' stored in memory 119. Signal processing unit 120 can evaluate the dispersion of the normalized second valley V12' obtained by the normalization process in step S14 of Fig. 16 by referring to the relationship between the dispersion and the normalized second valley V12' intensity stored in memory 119.

[0069] As described above, according to the second embodiment, it is possible to create a normalized intensity based on multiple valleys of the intensity of an optical signal transmitted through the optical fiber 131, and to determine whether the dispersion of the optical fiber is positive or negative based on the variation of the normalized intensity. Therefore, even for an installed optical fiber 131 whose zero-dispersion wavelength is unknown, it is possible to optimize performance by determining whether the dispersion is positive or negative.

[0070] Furthermore, according to the second embodiment, it is possible to estimate not only whether the dispersion of the optical fiber 131 is positive or negative but also the amount of dispersion by further using the known characteristics of the EA modulator 115. Since the dispersion of the optical fiber 131 can be determined by directly analyzing the PAM signal used in actual signal transmission, this method can be applied to optical transceivers.

[0071] (Modification) Next, a modification of the second embodiment will be described. Referring to the histogram shown on the left side of Fig. 17, the distribution of counts centered on the peak is generally a normal distribution using a Gaussian function because it is due to thermal noise and shot noise. Since the second embodiment assumes PAM4, the distribution function of the entire histogram can be assumed to be the sum of four Gaussian functions. In this modification, the variance is determined from the spread of the Gaussian function, which is the distribution function of the histogram, i.e., the standard deviation.

[0072] Fig. 19 is a diagram illustrating the latter half of the process in the signal processing unit 120 of the modified example. The modified example differs from the second embodiment in the latter half of the process in the signal processing unit 120 shown in Fig. 19. The other configurations are the same as those of the second embodiment.

[0073] In step S23, the signal processing unit 20 detects peaks in the histogram and calculates the discretized intensities of four peaks from the 0th level L0 to the 3rd level L3.

[0074] In step S24, normalization was performed so that the intensities of the 0th peak L0 and the 3rd peak L3 were −3 and +3, respectively. As described above, since it is assumed that the four signal levels of the PAM are equally spaced, the intensities of the 1st peak L1 and the 2nd peak L2 are −1 and +1, respectively, and the interval between the intensities of each peak after normalization is 2.

[0075] In step S25, the normalized intensity is fitted with a distribution function. The distribution function is a sum function of Gaussian functions with four independent peaks, as shown in Equation 1. In Equation 1, the independent variable x is the intensity, and the dependent variable y is the count number. The 0th count number N0, the first count number N1, the second count number N2, and the third count number N3 are the count numbers at the 0th peak L0, the first peak L1, the second peak L2, and the third peak L3, respectively. The 0th standard deviation σ0, the first standard deviation σ1, the second standard deviation σ2, and the third standard deviation σ3 are the standard deviations of the Gaussian functions whose axes are the 0th peak L0, the first peak L1, the second peak L2, and the third peak L3, respectively.

[0076] In fact, the normalized histogram intensities were linearly displayed to match the normalization and fitted with a distribution function. When the dispersion of the optical fiber 131 is 0, the zeroth standard deviation σ of the zeroth peak L is 0.22, but the first standard deviation σ of the first peak L, the second standard deviation σ of the second peak L, and the third standard deviation σ of the third peak are all approximately the same, 0.24.

[0077] 20 is a graph showing the relationship between the standard deviation and dispersion when the dispersion amount of the optical fiber 131 is changed. While there is no significant variation in the standard deviation in the region where the dispersion is from 0 to positive, in the region where the dispersion is negative, the standard deviation shows a clear tendency of σ0 < σ1 < σ2 < σ3. Therefore, based on the magnitude relationship of each standard deviation, if σ0 < σ1 < σ2 < σ3, it can be determined that the dispersion of the optical fiber 131 is negative.

[0078] In the modified example, the dispersion of the optical fiber 131 is determined based on the standard deviation found by fitting the histogram of light intensity to a distribution function. Because the standard deviation is found by fitting to a distribution function, whether the dispersion is positive or negative can be determined relatively accurately even if the histogram of the measured light intensity contains noise.

[0079] The signal processing unit 120 and memory 119 of the dispersion measurement device 110 described above can be, for example, a general-purpose computer system. The computer system includes a CPU (Central Processing Unit, or processor), memory, storage (HDD: Hard Disk Drive, SSD: Solid State Drive), a communication device, an input device, an output device, and the like. The memory and storage are storage devices. In this computer system, the CPU executes a predetermined program loaded into the memory, thereby realizing each function of the signal processing unit 120. The program for the signal processing unit 120 can be stored in a computer-readable recording medium such as an HDD, SSD, USB (Universal Serial Bus) memory, CD (Compact Disc), or DVD (Digital Versatile Disc), or can be distributed via a network. The computer-readable recording medium is, for example, a non-transitory recording medium.

[0080] It should be noted that the present disclosure is not limited to the second embodiment, and various modifications are possible within the scope of the gist thereof.

[0081] The present disclosure can be used to optimize performance in accordance with the transmission path in IMDD transmission of high-speed optical signals exceeding 25 Gbaud.

[0082] REFERENCE SIGNS LIST 10 Dispersion measuring device 11 DFB laser 12 PAM signal generator 13 DC bias voltage generator 14 AC / DC signal superimposing device 15 EA modulator 16 Optical amplifier 17 O / E converter 18 A / D converter 19 Memory 20 Signal processing unit 110 Dispersion measuring device 111 DFB laser 112 PAM signal generator 113 DC bias voltage generator 114 AC / DC signal superimposing device 115 EA modulator 116 Optical amplifier 117 O / E converter 118 A / D converter 119 Memory 120 Signal processing unit

Claims

1. A dispersion measurement device for measuring the dispersion of an optical fiber, comprising: a laser light source; an electro-absorption modulator that generates an optical signal by performing pulse amplitude modulation on laser light supplied from the laser light source; an opto-electrical converter that receives the optical signal that has propagated through the optical fiber; and a signal processing unit that determines whether the dispersion of the optical fiber is positive or negative based on a displacement of an inner eye level that makes the height of the inner eye opening asymmetric for an eye pattern of the optical signal received by the opto-electrical converter.

2. The dispersion measuring device of claim 1, wherein the signal processing unit determines that the dispersion of the optical fiber is negative when the level of the inner eye is displaced such that the height of the opening of the upper inner eye is greater than the height of the opening of the lower inner eye, and determines that the dispersion of the optical fiber is positive when the level of the inner eye is displaced such that the height of the opening of the lower inner eye is greater than the height of the opening of the upper inner eye.

3. The dispersion measuring device of claim 1, further comprising a memory storing a relationship between the inner eye level of the optical signal normalized by the outer eye level of the eye pattern and the dispersion of the optical fiber, wherein the signal processing unit normalizes the optical signal by the outer eye level for the eye pattern of the received optical signal, and estimates the dispersion of the optical fiber from the inner eye level of the normalized optical signal by referring to the relationship stored in the memory.

4. A dispersion measurement method for measuring the dispersion of an optical fiber, comprising: performing pulse amplitude modulation on laser light using an electroabsorption modulator to generate an optical signal; photoelectrically converting the optical signal propagating through the optical fiber; and determining whether the dispersion of the optical fiber is positive or negative based on a displacement of an inner eye level that makes the height of the opening of the inner eye asymmetric with respect to an eye pattern of the photoelectrically converted optical signal.

5. The dispersion measurement method described in claim 4, wherein the dispersion is determined to be negative when the level of the inner eye is displaced such that the height of the upper inner eye opening is greater than the height of the lower inner eye opening, and the dispersion is determined to be positive when the level of the inner eye is displaced such that the height of the lower inner eye opening is greater than the height of the upper inner eye opening.

6. The dispersion measurement method according to claim 4, further comprising: performing analog-to-digital conversion on the photoelectrically converted data; generating a histogram of the data discretized by the analog-to-digital conversion; determining outer eye and inner eye levels of the optical signal based on the histogram; and normalizing the inner eye level based on the outer eye level.

7. The dispersion measurement method according to claim 6, further comprising storing a relationship between the dispersion of said optical fibers having known different dispersions and the displacement of said inner eye level of said normalized optical signal.

8. The dispersion measurement method according to claim 7, further comprising estimating the dispersion of the optical fiber by referring to the stored relationship for the displacement of the inner eye level of the normalized optical signal.

9. A dispersion measurement device for measuring the dispersion of an optical fiber, comprising: a laser light source; an electro-absorption modulator for generating an optical signal by performing pulse amplitude modulation on laser light supplied from the laser light source; an opto-electrical converter for receiving the optical signal that has propagated through the optical fiber; and a signal processing unit for determining whether the dispersion of the optical fiber is positive or negative based on the optical intensity of multiple valleys in a histogram of optical intensity sampled from the optical signal received by the opto-electrical converter.

10. The dispersion measuring device of claim 9, wherein the signal processing unit determines the positive and negative of the normalized value obtained by normalizing the light intensity of the central valley by the light intensity of the lowest and highest valleys in the direction of light intensity as the positive and negative of the dispersion.

11. The dispersion measuring device according to claim 10, further comprising a memory storing a relationship between the normalized value and the dispersion of the optical fiber, wherein the signal processing unit estimates the dispersion of the optical fiber from the normalized value by referring to the relationship stored in the memory.

12. The dispersion measuring device according to claim 9, wherein the signal processing unit fits a Gaussian function to each of the multiple peaks of the histogram, and determines that the dispersion of the optical fiber is negative when the variance of the fitted Gaussian function increases as the peak moves upward in the light intensity direction.

13. A dispersion measurement method for measuring the dispersion of an optical fiber, comprising: performing pulse amplitude modulation on laser light using an electroabsorption modulator to obtain an optical signal; photoelectrically converting the optical signal propagated through the optical fiber; and determining whether the dispersion of the optical fiber is positive or negative based on the optical intensity of a plurality of valleys in a histogram of the optical intensity sampled from the photoelectrically converted optical signal.

14. The dispersion measuring method according to claim 13, wherein the level of the central valley is normalized to the levels of the lowest and highest valleys in the direction of light intensity.

15. The dispersion measuring method according to claim 14, wherein the positive and negative normalized values ​​correspond to the positive and negative dispersions.

16. The dispersion measuring method according to claim 14, further comprising storing a relationship between the dispersion of said optical fibers having known different dispersions and said normalized value.

17. The dispersion measuring method according to claim 16, further comprising estimating the dispersion of said optical fiber from said normalized value by referring to said stored relationship.

18. A dispersion measurement method as described in claim 13, in which a Gaussian function is fitted to each of a plurality of peaks formed in the histogram, and when the standard deviation of the fitted Gaussian function increases as the peaks move upward in the direction of light intensity, it is determined that the dispersion of the optical fiber is negative.

Images