JP5977211B2 - Optical transmission system and signal processing method - Google Patents

Description

  The present invention relates to an optical transmission system and a signal processing method for calculating a signal-to-noise power ratio in optical communication and improving a transmission method based on the calculated signal-to-noise power ratio.

  In optical communication, detecting the signal-to-noise power ratio (S / N) of an optical signal that can be realized on an optical transmission line is useful in the following two points. The first is that efficient transmission is possible in all paths. Theoretically, the maximum capacity that can be transmitted on each optical path is determined by the signal S / N. Therefore, it is possible to set a transmission method that realizes the maximum capacity on each optical path by detecting the signal S / N. Use limited resources as efficiently as possible. In addition, when the transmission capacity is fixed, the signal band is reduced by selecting the most efficient modulation method that can be realized by the signal S / N on each optical path and reducing the redundancy, thereby maximizing the resource efficiency. Can also be used. Second, it is possible to prevent the occurrence of errors. If the signal S / N falls below a certain threshold value due to optical fiber damage or the like, an error occurs in the optical path. it can.

  In general, there are two methods for measuring the signal S / N. The first method is a polarization nulling method, in which single-polarization transmission is performed, and the signal power of one polarization is compared with the noise power of a polarization orthogonal to the signal power. In this signal S / N measurement method, an optical signal is generated on a single polarization plane in a transmission unit, transmitted through an optical transmission line, and each polarization component is separated by a polarization separation circuit in a reception unit. The S / N of the signal after transmission can be obtained by comparing the power of the polarization component in which the noise exists and the power of the noise component of the polarization plane in the orthogonal relationship. The second method is a method of comparing the power of a signal component and the power of a noise component outside the signal band within the same polarization. This method is a method for obtaining the S / N of a signal from the power of the signal component of the received signal and the noise power information that can be measured from the gap between adjacent channels (Non-Patent Document 1).

Zhongqi Pan, Changyuan Yu, Alan E. Willner., “Optical performance monitoring for the next generation optical communication networks,” Optical Fiber Technology 16, pp.20-45, January 2010. <http://www.sciencedirect.com/ science / article / pii / S1068520009000686>

  Currently, in optical communication, high-speed 100 Gb / s transmission has become a reality, and the polarization multiplexing QPSK method is adopted as a transmission method. Therefore, it is fundamental to perform polarization multiplexing, and it is not practical to use the polarization nulling method. In the polarization nulling method, a signal is placed only on a single polarization, and the orthogonal polarization side is set as a noise measurement channel. In polarization multiplexing transmission, there is a problem that a noise component cannot be measured and the S / N of the signal cannot be calculated because a signal is also present on the orthogonal polarization.

  As a typical example of the second method described above, there is an Out of band method for measuring noise power from a band between adjacent signal channels. In this method, if the signal modulation speed increases and the signal band is expanded along with it, the frequency interval between adjacent channels becomes narrower and the density becomes higher. Therefore, it is difficult to secure a frequency band for measuring the power of the noise component. become. In a reconfigurable optical add / drop multiplexer (ROADM) used in a relay system, noise components outside a signal band are removed by an optical filter. Therefore, there is a problem that noise power cannot be measured from noise components outside the signal band. For example, a frequency grid of 50 GHz per channel is used in optical communication, but in recent 100 Gb / s transmission, the band exceeds 40 GHz, so noise components outside the signal band are removed by an optical filter. It will be.

  The present invention has been made to solve the above problems, and an object of the present invention is to provide an optical transmission system capable of obtaining a signal-to-noise power ratio regardless of transmission by polarization multiplexing or configuration in a transmission path in optical communication, and It is to provide a signal processing method.

  In order to solve the above problem, the present invention provides an optical transmission system that performs transmission using an optical signal in which a specific frequency signal sequence in which power is concentrated at a specific frequency and a transmission data sequence are time-division multiplexed. An FFT unit that converts a digital signal sequence obtained from an optical signal into a frequency domain signal, and a power value of a frequency at which the power of the specific frequency signal sequence in the frequency domain signal converted by the FFT unit is concentrated A section specifying unit for specifying a section in which the specific frequency signal sequence is arranged in the digital signal sequence; a signal extracting unit for extracting a digital signal sequence in the section specified by the section specifying unit from the digital signal sequence; A specific frequency output by removing the component of the specific frequency signal sequence from the digital signal sequence extracted by the signal extraction unit; Based on the digital signal sequence output from the specific frequency signal component removal unit, a noise power calculation unit that calculates noise power that is noise power contained in the received optical signal, and the noise power An optical transmission system comprising: a power ratio calculation unit that calculates a ratio between the power of a transmission data sequence and the power of noise based on the noise power calculated by the calculation unit.

  Further, the present invention provides the digital signal sequence obtained from the received optical signal by performing an FFT with a number of points smaller than the number of symbols of the specific frequency signal sequence. The frequency domain signal is converted into the frequency domain signal, or the section identification unit includes a section in the frequency domain signal in which power exceeding the threshold power is converted by the FFT unit, and the specific frequency signal sequence is included in the digital signal sequence. The FFT section specifies a digital signal sequence obtained from the received optical signal by performing FFT with a number of points smaller than the number of symbols of the specific frequency signal sequence. In the section specifying unit, power exceeding a threshold power is converted by the FFT unit. The section existing in the signal in the frequency domain, and identifies as a section in which the specific frequency signal sequence in the digital signal sequence is disposed.

  According to the present invention, in the above-described invention, a filter unit that extracts a signal in the vicinity of a frequency where the power of the specific frequency signal sequence is concentrated from the frequency domain signal converted by the FFT unit, and the filter unit includes: A first IFFT unit that converts the extracted signal into a time domain signal; a power calculation unit that calculates a power value of the time domain signal converted by the first IFFT unit for each symbol; and the power calculation unit Further comprising an averaging circuit that calculates a moving average value of the power value calculated by the interval specifying unit, based on a time at which the moving average value calculated by the averaging circuit is maximized, in the digital signal sequence A section in which the specific frequency signal sequence is arranged is specified.

  According to the present invention, in the above-described invention, the power is concentrated on a first specific frequency signal sequence in which power is concentrated on the specific frequency and on a frequency different from a frequency on which the power of the first specific frequency signal sequence is concentrated. And a specific frequency signal sequence generation unit that generates a signal sequence obtained by concatenating the second specific frequency signal sequence to be generated as the specific frequency signal sequence.

  The present invention further includes a second IFFT unit that converts a digital signal sequence used by the noise power calculation unit to calculate the noise power into a time domain signal. Is characterized in that the noise power is calculated based on a time-domain signal converted by the second IFFT unit.

  In the invention described in the above, the present invention shares the arithmetic function for performing IFFT in the first IFFT unit and the second IFFT unit in a time division manner, or sets the power value in the noise power calculation unit. A calculation function for calculating the power of the specific frequency signal sequence and the calculation of the noise power by sharing the calculation function to be calculated in time division, or performing an IFFT in the first IFFT unit and the second IFFT unit The function is shared in time division, and the calculation function for calculating the power value in the noise power calculation unit is shared in time division to calculate the power of the specific frequency signal sequence and the noise power. To do.

  Further, the present invention is the above-described invention, wherein the signal extraction unit extracts a digital signal sequence from a section other than the section specified by the section specifying section, or the noise power calculation section is the section specifying section The signal power, which is the power of the digital signal sequence including the transmission data sequence, is calculated based on a digital signal sequence other than the section specified by the power ratio, and the power ratio calculation unit calculates the transmission power from the signal power and the noise power. The ratio of the power of the data sequence and the power of the noise is calculated, or the signal extraction unit extracts a digital signal sequence from a section other than the section specified by the section specifying section, and the noise power calculation section A signal that is the power of a digital signal sequence including the transmission data sequence based on a digital signal sequence other than the interval specified by the interval specifying unit extracted by the extraction unit Calculating a force, the power ratio calculation unit, and calculates the ratio of the power and the noise power of the transmission data sequence from the said signal power and the noise power.

  The present invention further includes a low frequency band component removing unit that removes a low frequency band component from the digital signal sequence output from the specific frequency signal component removing unit and outputs the digital signal sequence. The noise power calculation unit calculates the noise power based on the digital signal sequence output from the low frequency band component removal unit.

  Further, the present invention provides the modulation method used for the signal sequence in which the specific frequency signal sequence and the transmission data sequence are time-division multiplexed in accordance with the ratio calculated by the power ratio calculation unit in the invention described above. A transmission format determination unit that determines at least one of a data rate, a signal band, or a redundancy, an optical signal transmission unit that encodes and modulates the transmission data sequence based on the determination of the transmission format determination unit, and And a demodulator that performs demodulation and decoding corresponding to the modulation scheme, data rate, signal band, and redundancy used in the optical signal transmitter.

  According to the present invention, in the invention described above, two optical signals corresponding to two different signal sequences in which the specific frequency signal sequence and the transmission data sequence are time-division multiplexed are multiplexed in a polarization state orthogonal to each other. A polarization multiplexing unit that outputs the received optical signal, and a polarization separation unit that separates the received optical signal into a first optical signal and a second optical signal whose polarization planes are orthogonal to each other, and the FFT unit includes: The digital signal sequence obtained from the first optical signal and the digital signal sequence obtained from the second optical signal are converted into frequency domain signals.

Further, the present invention is the above-described invention, wherein the signal sequence in which the envelope shape of the frequency spectrum of the transmission data sequence and the power concentrated on the specific frequency are in a proportional relationship is generated as the specific frequency signal sequence. And a specific frequency signal sequence generation unit.
In the invention described above, the present invention generates, as the specific frequency signal sequence, a signal sequence in which the power level of the frequency spectrum of the transmission data sequence is proportional to the density of the specific frequency interval. A specific frequency signal sequence generation unit is further provided.
Further, the present invention provides the specific frequency signal sequence for generating, as the specific frequency signal sequence, a signal sequence in which the power value concentrated on the specific frequency is a power value corresponding to the frequency characteristic in the own system. It further has a generating part.

  In order to solve the above problem, the present invention provides a signal in an optical transmission system that performs transmission using an optical signal in which a specific frequency signal sequence in which power is concentrated at a specific frequency and a transmission data sequence are time-division multiplexed. An FFT step for converting a digital signal sequence obtained from a received optical signal into a frequency domain signal, and the power of the specific frequency signal sequence in the frequency domain signal converted by the FFT step is concentrated. A section specifying step for specifying a section in which the specific frequency signal sequence is arranged in the digital signal sequence based on a power value of a frequency to be transmitted, and a digital signal sequence in the section specified by the section specifying step from the digital signal sequence A signal extracting step for extracting the digital signal and a digital signal extracted in the signal extracting step Included in the received optical signal based on the digital signal sequence output in the specific frequency signal component removal step for removing the specific frequency signal sequence component from the signal signal sequence and outputting the signal. A noise power calculation step for calculating noise power, which is noise power, and a power ratio calculation step for calculating a ratio between the power of the transmission data sequence and the noise power based on the noise power calculated in the noise power calculation step; A signal processing method characterized by comprising:

According to the present invention, by specifying the section of the specific frequency signal sequence in the digital signal sequence obtained from the received optical signal based on the frequency at which the power of the specific frequency signal sequence is concentrated, the components of the transmission data sequence are determined. A digital signal sequence not included can be obtained.
By converting a digital signal sequence that does not include this transmission data sequence component into a frequency domain signal and removing the frequency component of the specific frequency signal sequence, a noise component can be obtained. The power ratio, i.e. the signal to noise power ratio, can be obtained.

It is a block diagram which shows the structural example of the optical transmission system in 1st Embodiment. It is a block diagram which shows the structural example of the optical signal transmitter 1 in the embodiment. It is a figure which shows an example of the frequency spectrum of the transmission data series produced | generated in the transmission data series production | generation circuits 13-1 and 13-2 in the embodiment. It is a figure which shows an example of the frequency spectrum of the specific frequency signal series which the specific frequency signal series generation circuit 14-1 in the embodiment produces | generates. It is a block diagram which shows the structural example of the optical signal receiver 3 in the embodiment. It is a block diagram which shows the structural example of the SN ratio calculation apparatus 4 in the embodiment. It is a figure which shows the outline | summary of FFT which the overlap FFT calculating circuit 42-1 in the embodiment performs. It is a figure which shows an example of the frequency spectrum of the digital signal series input from the optical signal receiver 3 to the SN ratio calculation apparatus 4 in the same embodiment. FIG. 9 is a conceptual diagram obtained by converting the digital signal series shown in FIG. 8 with time on the horizontal axis and power on the vertical axis. It is a conceptual diagram which shows the example of a result of the signal processing by the training signal extraction circuit 44-1 in the same embodiment. It is a conceptual diagram which shows the example of a result of the signal processing by the overlap FFT calculating circuit 45-1 in the same embodiment. It is a figure which shows the outline | summary of the determination of the area used for calculation of noise power in the embodiment. It is a block diagram which shows the structural example of 4 A of SN ratio calculation apparatuses in 2nd Embodiment. It is a figure which shows an example of the frequency spectrum output from the frequency bandpass filter circuit 404-11 in the same embodiment. It is a figure which shows an example of the signal of the time domain input into the averaging circuit 408-1 to 408-N in the embodiment. It is a figure which shows an example of the moving average value of the electric power which the averaging circuits 408-1 to 408-N in the embodiment output. It is a figure which shows an example of the specific frequency signal sequence using the 1st specific frequency signal sequence and the 2nd specific frequency signal sequence. It is a block diagram which shows the structural example of SN ratio calculation apparatus 4B in 3rd Embodiment. It is a block diagram which shows the structural example of the training signal area detection block 40 in the embodiment. It is a block diagram which shows the structural example of the power ratio calculation block 60 in the embodiment. It is a figure which shows the spectrum example of the digital signal series input into the signal extraction circuits 62-1 and 62-2 at the time of non-transmission. It is a figure which shows the example of a spectrum of the training signal component extracted at the time of non-transmission. It is a figure which shows the spectrum of the specific frequency signal series after transmission measured in experiment. It is a figure which shows the structure of a Mach-Zehnder type | mold optical modulator. It is a flowchart which shows the SN ratio calculation process by the SN ratio calculation apparatus 4B in the same embodiment. It is a figure which shows the effect of the circuit scale reduction in the SN ratio calculation apparatus 4B of the embodiment. It is a figure which shows the example of the specific frequency signal sequence from which the frequency spectrum of a transmission data sequence and a specific frequency signal sequence becomes a proportional relationship. It is a figure which shows an example of the frequency spectrum of the specific frequency signal sequence in which the density of the space | interval of the frequency where the electric power of a specific frequency signal sequence concentrates, and the power level of a transmission data sequence have a proportional relationship. It is a figure which shows an example of the frequency spectrum of the specific frequency signal sequence from which the power level in each frequency where the electric power of a specific frequency signal sequence concentrates becomes proportional to the loss factor of the frequency characteristic in an optical transmission system.

  Hereinafter, an optical transmission system and a signal processing method according to embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram illustrating a configuration example of an optical transmission system according to the first embodiment. The optical transmission system according to this embodiment includes an optical signal transmission device 1, an optical transmission path 2, an optical signal reception device 3, an SN ratio calculation device 4, and a control plane 5. The optical signal transmitter 1 converts transmission data to be transmitted into an optical signal. The optical signal transmission device 1 transmits the optical signal obtained by the conversion to the optical signal reception device 3 through the optical transmission path 2. The optical signal receiver 3 receives the optical signal transmitted from the optical signal transmitter 1. The optical signal receiving device 3 converts the received optical signal into an electrical signal, and acquires transmission data from the electrical signal. Further, the optical signal receiving device 3 outputs an electrical signal obtained by converting the optical signal to the SN ratio calculating device 4.

  The SN ratio calculation device 4 calculates the signal-to-noise power ratio (SN ratio) of the optical signal in the optical transmission system based on the electrical signal output from the optical signal receiving device 3. This SN ratio is a value related to transmission data. The SN ratio calculation device 4 outputs an SN ratio signal indicating the calculated SN ratio to the control plane 5. The control plane 5 controls the transmission amount and the like according to the S / N ratio indicated by the S / N ratio signal. Further, the control plane 5 outputs the SN ratio signal output from the SN ratio calculation device 4 to the optical signal transmission device 1 and the optical signal reception device 3. The optical signal transmission device 1 and the optical signal reception device 3 are used in accordance with the S / N ratio indicated by the S / N ratio signal, the modulation method, data rate, and signal band used between the optical signal transmission device 1 and the optical signal reception device 3. And redundancy in forward error correction (FEC) is set.

  FIG. 1 shows an example in which the optical transmission system includes one optical signal transmission device 1 and one optical signal reception device 3. However, the optical signal transmission device 1 and the optical signal reception device 3 may have a plurality of configurations. In this case, when a plurality of sets of optical signal transmitters 1 and optical signal receivers 3 can be used for transmission of transmission data, the control plane 5 uses the optical signal transmitters 1 and optical signals used for transmitting transmission data. You may make it determine the number with the receiver 3 according to SN ratio. Further, when the SN ratio calculation device 4 outputs the SN ratio to the control plane 5, the SN ratio may be transmitted by either an electric signal or an optical signal. In addition, although the optical transmission system illustrated in FIG. 1 includes the control plane 5, the optical transmission system may not include the control plane 5. In this case, the SN ratio signal may be output to the optical signal receiving device 3 and further output from the optical signal receiving device 3 to the optical signal transmitting device 1.

  FIG. 2 is a block diagram illustrating a configuration example of the optical signal transmission device 1 according to the present embodiment. The optical signal transmitter 1 includes an SN ratio acquisition circuit 11, transmission format determination circuits 12-1 and 12-2, transmission data sequence generation circuits 13-1 and 13-2, and specific frequency signal sequence generation circuits 14-1 and 14-. 2, signal multiplexing circuits 15-1 and 15-2, electrical / optical conversion circuits 16-1 and 16-2, and a polarization multiplexing circuit 17. The optical signal transmitter 1 includes a configuration from the transmission format determination circuit 12-1 to the electrical / optical conversion circuit 16-1, and a configuration from the transmission format determination circuit 12-2 to the electrical / optical conversion circuit 16-2. There are two blocks. Two optical signals generated in each block are polarization multiplexed and transmitted.

  In the following description, an optical signal having one polarization among orthogonal polarizations will be described as a first optical signal, and an optical signal having the other polarization will be described as a second optical signal. Further, the configuration from the transmission format determining circuit 12-1 to the electrical / optical conversion circuit 16-1 for generating the first optical signal, and the transmission format determining circuit 12-2 for generating the second optical signal. To the electrical / optical conversion circuit 16-2, the configuration corresponding to the first optical signal will be described, and the description of the configuration corresponding to the second optical signal will be omitted.

  The S / N ratio acquisition circuit 11 receives the S / N ratio signal output via the control plane 5, and outputs the S / N ratio indicated by the received S / N ratio signal to each of the two transmission format determination circuits 12-1 and 12-2. . The transmission format determination circuit 12-1 is based on the S / N ratio output from the S / N ratio acquisition circuit 11 and uses the modulation method, data rate, signal band, and FEC redundancy used when transmitting transmission data to the optical signal receiver 3. (Coding rate) is determined. For example, the transmission format determination circuit 12-1 stores in advance a table in which the modulation scheme, data rate, signal band, and FEC redundancy are associated with the SN ratio, and is output from the SN ratio acquisition circuit 11. It is determined to use the modulation scheme, data rate, signal band, and FEC redundancy corresponding to the S / N ratio. The transmission format determination circuit 12-1 outputs a combination of the determined modulation scheme, data rate, signal band, and FEC redundancy to the transmission data sequence generation circuit 13-1.

  Alternatively, the transmission format determination circuit 12-1 may determine a modulation scheme that provides a preset code error rate from the SN ratio per bit. For example, when the SNR per bit is specified to be QPSK when the modulation scheme that matches a preset code error rate is 15 dB, if the SNR per bit is 12 dB, the BPSK is set to 19 dB. If so, it is determined to be 16QAM.

  For example, when the S / N ratio obtained via the S / N ratio calculation device 4 has a higher value than the predetermined requirement for the optical transmission system, the transmission format determination circuit 12-1 selects the modulation method. Switch to a high multilevel method, increase the data rate, or lower the FEC redundancy. That is, when the SN ratio is equal to or greater than the predetermined requirement, the transmission format determination circuit 12-1 increases the data amount per unit time of transmission data transmitted from the optical signal transmission device 1 to the optical signal reception device 3.

  Similarly, when the S / N ratio obtained via the S / N ratio calculation device 4 has a low value as compared with the requirements predetermined for the optical transmission system, the transmission format determination circuit 12-1 selects the modulation method. Are switched to a low multi-value scheme, the data rate is lowered, or the FEC redundancy is increased. That is, when the SN ratio is less than or equal to the predetermined requirement, the transmission format determination circuit 12-1 reduces the data amount per unit time of transmission data transmitted from the optical signal transmission device 1 to the optical signal reception device 3.

  In addition, when the SN ratio obtained via the SN ratio calculation device 4 is a value higher than the requirement predetermined for the optical transmission system, the transmission format determination circuit 12-1 has a transmission capacity of As it is, the modulation scheme can be switched to a scheme with a high multilevel or the FEC redundancy can be lowered. That is, when the SN ratio is equal to or greater than the predetermined requirement, the transmission format determination circuit 12-1 reduces the signal band of transmission data transmitted from the optical signal transmission device 1 to the optical signal reception device 3.

  In addition, when the SN ratio obtained via the SN ratio calculation device 4 is a low value compared to the requirements predetermined for the optical transmission system, the transmission format determination circuit 12-1 has a transmission capacity of As it is, the modulation scheme can be switched to a scheme with a low multilevel, or the FEC redundancy can be increased. That is, when the SN ratio is less than or equal to the predetermined requirement, the transmission format determination circuit 12-1 increases the signal band of transmission data transmitted from the optical signal transmission device 1 to the optical signal reception device 3.

  The transmission data sequence generation circuit 13-1 transmits data to be transmitted to the optical signal receiving device 3 based on the modulation method, data rate, signal band, and FEC redundancy determined by the transmission format determination circuit 12-1. A transmission data sequence is generated by encoding and modulation. Transmission data sequence generation circuit 13-1 outputs the generated transmission data sequence to signal multiplexing circuit 15-1. The specific frequency signal sequence generation circuit 14-1 generates a specific frequency signal sequence (training signal sequence) having power concentrated on a predetermined frequency component. The specific frequency signal sequence generation circuit 14-1 outputs the generated specific frequency signal sequence to the signal multiplexing circuit 15-1. Based on the power of the specific frequency signal sequence, the optical signal receiving device 3 calculates the SN ratio.

  The signal multiplexing circuit 15-1 receives the transmission data sequence generated by the transmission data sequence generation circuit 13-1 and the specific frequency signal sequence generated by the specific frequency signal sequence generation circuit 14-1. The signal multiplexing circuit 15-1 time-division multiplexes the transmission data sequence and the specific frequency signal sequence to generate a signal sequence in which the specific frequency signal sequence periodically appears between the transmission data sequences. The signal multiplexing circuit 15-1 outputs the generated signal series to the electrical / optical conversion circuit 16-1. The electrical / optical conversion circuit 16-1 converts the signal series output from the signal multiplexing circuit 15-1 into an optical signal and outputs the obtained optical signal to the polarization multiplexing circuit 17.

  The polarization multiplexing circuit 17 receives optical signals output from the two electric / optical conversion circuits 16-1 and 16-2. The polarization multiplexing circuit 17 multiplexes and outputs two input optical signals in a polarization state orthogonal to each other. For the polarization multiplexing circuit 17, for example, a polarization beam combiner is used. The optical signal output from the polarization multiplexing circuit 17 is transmitted to the optical signal receiving device 3 through the optical transmission path 2.

Here, signal sequences generated in the transmission data sequence generation circuits 13-1 and 13-2 and the specific frequency signal sequence generation circuits 14-1 and 14-2 will be described.
FIG. 3 is a diagram illustrating an example of a frequency spectrum of transmission data sequences generated in the transmission data sequence generation circuits 13-1 and 13-2 in the present embodiment. In the figure, the horizontal axis represents frequency and the vertical axis represents power. Here, the frequency spectrum is shown in the other case where the bandwidth of the transmission data sequence is 50 GHz.

  FIG. 4 is a diagram illustrating an example of a frequency spectrum of a specific frequency signal sequence generated by the specific frequency signal sequence generation circuit 14-1 in the present embodiment. In the figure, the horizontal axis represents frequency and the vertical axis represents power. Here, a frequency spectrum of a specific frequency signal sequence in which power is concentrated on two frequency components (± 12.5 GHz) is shown. The specific frequency signal sequence generation circuit 14-1 generates an alternating signal whose amplitude is repeated as (+ S, −S, + S, −S,...) As a specific frequency signal sequence, thereby generating a frequency spectrum as shown in FIG. It is possible to obtain a specific frequency signal sequence. Since this specific frequency signal series is amplitude-modulated and has a period of 2 samples, when the signal band is ± 25 GHz from the carrier frequency, a double-sideband signal is generated at ± 12.5 GHz as the power peak component of the specific frequency. Is done.

  The signal multiplexing circuits 15-1 and 15-2 time division multiplex the transmission data series as shown in FIG. 3 and the specific frequency signal series as shown in FIG. That is, an optical signal including a frequency spectrum obtained by superimposing the frequency spectrum shown in FIG. 3 and the frequency spectrum shown in FIG. 4 is output from the optical signal transmitting apparatus 1.

  Further supplementally, the specific frequency signal series can be generated by repeating alternating signals such as (0011) and (0101) when there are two frequencies where power is concentrated. For example, if a specific frequency signal sequence having an alternating pattern of (0101) is used for a 32 GBaud symbol, power is concentrated at a frequency that is ± 16 GHz away from the center frequency of the carrier wave. In addition, since the specific frequency signal sequence is time-division multiplexed with respect to the transmission data sequence, there is no transmission data sequence component at the time when the power peak of the specific frequency signal sequence appears. There is no interference. On the other hand, if the specific frequency signal sequence and the transmission data sequence are multiplexed by frequency multiplexing, interference occurs because respective components appear simultaneously. In this way, the specific frequency signal sequence generated by the specific frequency signal sequence generation circuit 14-1 can be generated at an arbitrary frequency, and can be prevented from interfering with the transmission data sequence.

  FIG. 5 is a block diagram illustrating a configuration example of the optical signal receiving device 3 in the present embodiment. The optical signal receiving apparatus 3 includes a polarization split circuit 31, optical / electrical conversion circuits 32-1 and 32-2, analog / digital conversion circuits (A / D conversion circuits) 33-1 and 33-2, and an adaptive equalization circuit. 34, and demodulation circuits 35-1 and 35-2. The optical signal receiving device 3 is provided with two optical / electrical conversion circuits 32, analog / digital conversion circuits 33, and demodulation circuits 35, which are orthogonal to each other included in the optical signal output from the optical signal transmission device 1. This is because processing is performed on signals corresponding to the two polarized waves.

The optical signal receiving device 3 includes a first system of an optical / electrical conversion circuit 32-1, an analog / digital conversion circuit 33-1, and a demodulation circuit 35-1, an optical / electrical conversion circuit 32-2, an analog / digital conversion circuit 32-1, There are blocks of the second system of the digital conversion circuit 33-2 and the demodulation circuit 35-2.
Each system block corresponds to two optical signals transmitted by polarization multiplexing. In addition, since the structure in the block of each system | strain is the same, it demonstrates with respect to the block of a 1st system | strain, and abbreviate | omits description with respect to the block of a 2nd system | strain.

  The polarization split circuit 31 receives an optical signal transmitted via the optical transmission path 2. The polarization split circuit 31 separates an optical signal included in the received optical signal into a first optical signal and a second optical signal whose planes of polarization are orthogonal to each other. The polarization splitting circuit 31 outputs the first optical signal to the optical / electrical conversion circuit 32-1, and outputs the second optical signal to the optical / electrical conversion circuit 32-2. The optical / electrical conversion circuit 32-1 converts the optical signal output from the polarization splitting circuit 31 into an analog electrical signal and outputs the analog electrical signal to the analog / digital conversion circuit 33-1.

  The analog / digital conversion circuit 33-1 converts the electrical signal output from the optical / electrical conversion circuit 32-1 into a first digital signal sequence digitized at a predetermined sampling frequency. The analog / digital conversion circuit 33-1 outputs the first digital signal sequence obtained by the conversion to the adaptive equalization circuit 34 and the SN ratio calculation device 4. The adaptive equalization circuit 34 includes two signals, a first digital signal sequence output from the analog / digital conversion circuit 33-1 and a second digital signal sequence output from the analog / digital conversion circuit 33-2. A digital signal sequence is input. The adaptive equalization circuit 34 compensates the first digital signal sequence for chromatic dispersion, polarization mode difference, phase noise, frequency characteristic distortion in the optical transmission system, and the like that occur in the optical transmission line 2 and the like. The adaptive equalization circuit 34 may perform compensation based on the S / N ratio signal. The adaptive equalization circuit 34 outputs a digital signal sequence obtained by compensation for the first digital signal sequence to the demodulation circuit 35-1. The adaptive equalization circuit 34 also performs compensation for the second digital signal sequence, and outputs the digital signal sequence obtained by the compensation to the demodulation circuit 35-2.

  The demodulation circuit 35-1 demodulates and decodes the digital signal sequence output from the adaptive equalization circuit 34, and acquires transmission data. Similarly, the demodulation circuit 35-2 also demodulates and decodes the digital signal sequence output from the adaptive equalization circuit 34, and acquires transmission data. The demodulating circuits 35-1 and 35-2 output the acquired transmission data to a data processing device connected to the optical signal receiving device 3. Note that the demodulation circuits 35-1 and 35-2 perform demodulation and decoding according to the modulation scheme, data rate, signal band, and FEC redundancy determined by the transmission format determination circuit 12-1.

  FIG. 6 is a block diagram showing a configuration example of the SN ratio calculation apparatus 4 in the present embodiment. The SN ratio calculation device 4 receives the first digital signal sequence from the analog / digital conversion circuit 33-1 of the optical signal reception device 3, and receives the second digital signal sequence from the analog / digital conversion circuit 33-2. Is done. The SN ratio calculation device 4 includes signal distribution circuits 41-1 and 41-2, overlap FFT calculation circuits 42-1 and 42-2, a peak power time calculation circuit 43, training signal extraction circuits 44-1 and 44-2, Overlap FFT circuits 45-1 and 45-2, a multiplexing circuit 46, a signal power calculation circuit 47, a specific frequency signal component removal circuit 48, a noise power calculation circuit 49, and an SN ratio calculation circuit 50 are provided.

  In the following description, the signal distribution circuit 41-1, the overlap FFT operation circuit 42-1, the training signal extraction circuit 44-1, and the overlap FFT operation circuit 45-1 perform signal processing on the first digital signal sequence. The signal distribution circuit 41-2, the overlap FFT operation circuit 42-2, the training signal extraction circuit 44-2, and the overlap FFT operation circuit 45-2 perform signal processing on the second digital signal sequence. explain. Each of the signal distribution circuit 41-1, the overlap FFT operation circuit 42-1, the training signal extraction circuit 44-1, and the overlap FFT operation circuit 45-1, the signal distribution circuit 41-2, and the overlap FFT operation Since the circuit 42-2, the training signal extraction circuit 44-2, and the overlap FFT operation circuit 45-2 have the same configuration, each circuit that performs signal processing on the second digital signal sequence will be described. Is omitted. Note that the frequency at which the power of the specific frequency signal sequence is concentrated in the SN ratio calculation device 4 is known.

  The first digital signal sequence is input from the optical signal receiving device 3 to the signal distribution circuit 41-1. The signal distribution circuit 41-1 duplicates the input first digital signal sequence into two. The signal distribution circuit 41-1 outputs one of the two first digital signals to the overlap FFT operation circuit 42-1, and outputs the other first digital signal sequence to the training signal. The data is output to the extraction circuit 44-1.

  The overlap FFT operation circuit 42-1 performs FFT on the first digital signal sequence output from the signal distribution circuit 41-1, thereby converting the time domain signal to the frequency domain signal. At this time, the overlap FFT operation circuit 42-1 cuts out the digital signal to be subjected to the FFT from the first digital signal series so that the digital signal targeted in the previous FFT partially overlaps. In other words, processing is performed in which the FFT targets overlap. The resolution can be increased by the overlap processing for overlapping the FFT target.

  Further, the number of FFT points (window size) in the overlap FFT operation circuit 42-1 is set smaller than the number of symbols of the specific frequency signal sequence (training signal sequence) generated in the specific frequency signal sequence generation circuit 14-1. Has been. The overlap FFT operation circuit 42-1 performs FTT on the first digital signal sequence that is sequentially input, and the section of the first digital signal sequence subjected to FFT and the frequency spectrum obtained by the FFT ( Three-dimensional data including frequency and power). The overlap FFT operation circuit 42-1 outputs the acquired three-dimensional data to the peak power time calculation circuit 43. Note that the start time of the section may be included in the three-dimensional data instead of the section targeted for FFT. Further, the number of FFT points in the overlap FFT operation circuit 42-1 may be an integral multiple of the period length (number of samples) of the specific frequency signal sequence.

  FIG. 7 is a diagram showing an outline of FFT performed by the overlap FFT operation circuit 42-1 in the present embodiment. As shown in the figure, the overlap FFT operation circuit 42-1 cuts out a digital signal to be subjected to FFT in a window having a size shorter than the length of the specific frequency signal sequence, and performs FFT. At this time, the digital signal cut out by the overlap FFT operation circuit 42-1 is cut out so as to overlap (overlap) the target digital signal in the preceding and succeeding FFTs. As a result, when the FFT targets only the digital signal of the specific frequency signal sequence, a signal in which power is concentrated on the frequency component determined for the specific frequency signal sequence is obtained. The section of the window where the power of the frequency component corresponding to the specific frequency signal sequence is high is determined as the section where the specific frequency signal sequence in the first digital signal sequence is arranged.

  In the FFT performed by the overlap FFT operation circuit 42-1, the higher the overlap rate of the digital signal that is the object of the FFT, the higher the resolution on the time axis and the detection of the section in which the specific frequency signal sequence is arranged. Increases accuracy. Further, as the overlap rate is higher, the ratio of the transmission data sequence component to the frequency peak component is increased, and the detection accuracy of the section in which the specific frequency signal sequence is arranged increases. In particular, when the FFT window size is equal to the length of the specific frequency signal sequence, if the digital signal sequence to be FFT and the specific frequency signal sequence match, power is maximized in the frequency component defined for the specific frequency signal sequence. Therefore, it becomes easy to detect the section where the specific frequency signal sequence is arranged. Note that if the FFT window size is longer than the length of the specific frequency signal sequence, only the specific frequency signal sequence cannot be subjected to the FFT, and a digital signal of the transmission data sequence is included. Leaks, and the ratio between the transmission data sequence component and the frequency peak component becomes small. Therefore, it is difficult to detect a section where the specific frequency signal sequence is arranged.

Returning to FIG. 6, the description of the configuration of the SN ratio calculation device 4 will be continued.
The peak power time calculation circuit 43 includes a three-dimensional data section in which the power of the frequency component corresponding to the specific frequency signal sequence is the largest based on the three-dimensional data output from the overlap FFT calculation circuit 42-1. A first interval is detected. The first section is a section in which the specific frequency signal sequence is included in the first digital signal sequence. That is, the peak power time calculation circuit 43 specifies the first section including the specific frequency signal sequence (training signal sequence) based on the three-dimensional data. The peak power time calculation circuit 43 outputs the first section information indicating the identified first section to the training signal extraction circuit 44-2. Similarly, the peak power time calculation circuit 43 calculates the second interval in which the specific frequency signal sequence is included in the second digital signal sequence based on the three-dimensional data output from the overlap FFT operation circuit 42-2. Identify. Further, the peak power time calculation circuit 43 outputs second section information indicating the specified second section to the training signal extraction circuit 44-2.

  The section corresponding to the specific frequency signal sequence is identified by, for example, three-dimensionally obtained sequentially by setting the power concentrated on the frequency component corresponding to the specific frequency signal sequence to K times or more of the average power of the entire band. This can be done by comparing the power contained in the data. Specifically, in the specific frequency signal sequence generation circuits 14-1 and 14-2, the frequency spectrum of the entire frequency band of the specific frequency signal sequence (training signal sequence) is acquired, and all the power at each frequency is added and sampled. The average power per sample is calculated by dividing by the number. The specific frequency signal sequence generation circuits 14-1 and 14-2 generate a specific frequency signal sequence in which the power concentrated on the specific frequency component is K times the average power per sample. K is an arbitrary positive real number. By generating the specific frequency signal sequence in this way, the peak power time calculation circuit 43 can specify the section corresponding to the specific frequency signal sequence in the first digital signal sequence. Similarly, the peak power time calculation circuit 43 can specify the section corresponding to the specific frequency signal sequence in the second digital signal sequence.

  The training signal extraction circuit 44-1 receives the first digital signal sequence output from the signal distribution circuit 41-1 and the first interval information for the first digital signal sequence from the peak power time calculation circuit 43. The Based on the first section information, the training signal extraction circuit 44-1 extracts a digital signal series in a section in which the specific frequency signal series is included in the first digital signal series. The training signal extraction circuit 44-1 outputs the extracted digital signal series to the overlap FFT operation circuit 45-1.

  The overlap FFT operation circuit 45-1 performs an FFT on the digital signal sequence output from the training signal extraction circuit 44-1, thereby converting the time domain signal to the frequency domain signal. Similar to the overlap FFT operation circuit 42-1, the number of FFT points in the overlap FFT operation circuit 45-1 is set smaller than the number of symbols of the specific frequency signal sequence (training signal sequence). Similarly to the overlap FFT operation circuit 42-1, the overlap FFT operation circuit 45-1 converts the digital signal to be subjected to FFT from the digital signal sequence so that the digital signal targeted in the previous FFT is included. cut. The overlap FFT operation circuit 45-1 outputs a frequency domain signal (frequency spectrum) obtained by the FFT to the multiplexing circuit 46.

  The frequency circuit signal obtained in the overlap FFT operation circuit 45-1 and the frequency domain signal obtained in the overlap FFT operation circuit 45-2 are input to the multiplexing circuit 46. Since the digital signal corresponding to the specific frequency signal sequence (training signal sequence) is extracted in the training signal extraction circuits 44-1 and 44-2 from the frequency domain signal input to the multiplexing circuit 46, the transmission data The signal does not include the series. That is, the input frequency domain signal includes a frequency component of a specific frequency signal sequence and a frequency component of noise. The multiplexing circuit 46 multiplexes the frequency domain signals sequentially output from the overlap FFT arithmetic circuit 45-1 and the frequency domain signals sequentially output from the overlap FFT arithmetic circuit 45-2. The multiplexing circuit 46 outputs the signal obtained by the multiplexing to the signal power calculation circuit 47 and the specific frequency signal component removal circuit 48.

Based on the signal output from the multiplexing circuit 46, the signal power calculation circuit 47 calculates signal power that is the power of the specific frequency signal series. The signal power calculation circuit 47 outputs the calculated power value of the signal power to the SN ratio calculation circuit 50.
The specific frequency signal component removal circuit 48 removes the component of the specific frequency signal sequence from the signal output from the multiplexing circuit 46, and outputs the signal from which the component of the specific frequency signal sequence is removed to the noise power calculation circuit 49. The removal of the components of the specific frequency signal sequence in the specific frequency signal component removal circuit 48 can be performed by using a band pass filter or the like since the frequency at which the power of the specific frequency signal sequence is concentrated is known. The signal processing in the specific frequency signal component removal circuit 48 may be performed in the frequency domain or in the time domain. The noise power calculation circuit 49 calculates noise power, which is noise power, based on the signal output from the specific frequency signal component removal circuit 48. The noise power calculation circuit 49 outputs the calculated noise power value to the SN ratio calculation circuit 50.

  Based on the noise power calculated by the noise power calculation circuit 49 and the signal power calculated by the signal power calculation circuit 47, the SN ratio calculation circuit 50 calculates an SN ratio as a ratio of the power of the transmission data sequence to the noise power. To do. The SN ratio calculation circuit 50 outputs an SN ratio signal indicating the calculated SN ratio to the optical signal transmitter 1, the optical signal receiver 3, and the control plane 5.

  The SN ratio calculation circuit 50 calculates the SN ratio of the transmission data sequence by calculating the SN ratio of the specific frequency signal sequence and then performing conversion. As a conversion method, for example, the power ratio per unit time between the specific frequency signal sequence and the transmission data sequence is used as a conversion factor, and the SN ratio of the transmission data sequence is calculated by multiplying the SN ratio of the specific frequency signal sequence by the conversion factor. calculate. For example, the conversion coefficient is derived in advance by comparing the power when the transmission data sequence and the specific frequency signal sequence are time-division multiplexed. Alternatively, in the simulation, the conversion coefficient may be derived based on the ratio between the transmission data sequence length and the specific frequency signal sequence length. Or, by comparing the SN ratio with high accuracy measured using a spectrum analyzer or the like with the SN ratio calculated by the SN ratio calculation device 4, and performing calibration for the SN ratio calculated by the SN ratio calculation device 4. A conversion coefficient may be determined.

  The SN ratio calculation circuit 50 calculates the SN ratio using the power of the first digital signal sequence and the second digital signal sequence input from the optical signal receiving device 3 instead of the power of the specific frequency signal sequence. You may make it do. Alternatively, a digital signal including components of the transmission data sequence is extracted from the first digital signal sequence and the second digital signal sequence based on the section specified by the peak power time calculation circuit 43, and the transmission data sequence is extracted from the extracted digital signal. The S / N ratio may be calculated using the power of the transmission data sequence instead of the power of the specific frequency signal sequence. In this case, since the power from which the power of the specific frequency signal sequence is removed is used, the accuracy can be improved as compared with the case where the power of the first digital signal sequence and the second digital signal sequence is used.

  Furthermore, the power of the first digital signal sequence and the second digital signal sequence can be obtained using the power of the optical signal incident on the optical signal receiving device 3. Since the transmission data sequence occupies most of the power ratio between the transmission data sequence and the specific frequency signal sequence in the optical signal received by the optical signal receiving device 3, for example, the optical / electrical conversion circuits 32-1 and 32-2. May be regarded as an optical signal incident on the. The signal power can be obtained by subtracting the noise power calculated in the noise power calculation circuit 49 from the obtained optical signal power.

Here, the signal processing in the SN ratio calculation apparatus 4 will be further described.
FIG. 8 is a diagram illustrating an example of a frequency spectrum of a digital signal sequence input from the optical signal receiving device 3 to the SN ratio calculating device 4 in the present embodiment. In the figure, the horizontal axis represents frequency and the vertical axis represents power. The frequency spectrum shown in the figure is transmitted by time-division multiplexing the transmission data sequence corresponding to the frequency spectrum shown in FIG. 3 and the specific frequency signal sequence corresponding to the frequency spectrum shown in FIG. Corresponds to the frequency spectrum of the signal. As shown in the figure, the frequency spectrum includes a frequency component of the transmission data sequence, a frequency component of the specific frequency signal sequence, and a frequency component of noise.

  FIG. 9 is a conceptual diagram obtained by converting the digital signal sequence shown in FIG. 8 with time on the horizontal axis and power on the vertical axis. As shown in the figure, when the digital signal sequence is viewed in the time axis direction, the specific frequency signal sequence (training signal sequence) is periodically time-division multiplexed with respect to the transmission data sequence and includes noise. Yes. In the S / N ratio calculation device 4, a signal as shown in FIG. 9 is distributed by the signal distribution circuit 41-1 to the overlap FFT calculation circuit 42-1 and the training signal extraction circuit 44-1.

  In the peak power time calculation circuit 43, as described above, the section including the component of the specific frequency signal sequence is specified based on the three-dimensional data output from the overlap FFT calculation circuit 42-1. In the training signal extraction circuit 44-1, a signal as shown in FIG. 10 is extracted from the signal as shown in FIG. 10 by extracting a signal based on the section of the specific frequency signal sequence specified by the peak power time calculation circuit 43. Is obtained.

  FIG. 10 is a conceptual diagram showing a result example of signal processing by the training signal extraction circuit 44-1 in the present embodiment. In the figure, the horizontal axis indicates time and the vertical axis indicates power. As shown in the figure, the signal extracted by the training signal extraction circuit 44-1 is a signal including a specific frequency signal sequence and noise in a section where the specific frequency signal sequence exists, with components of the transmission data sequence being removed. It becomes. The signal extracted by the training signal extraction circuit 44-1 is converted into a frequency spectrum as shown in FIG. 11 by the overlap FFT operation circuit 45-1.

  FIG. 11 is a conceptual diagram showing a result example of signal processing by the overlap FFT operation circuit 45-1 in the present embodiment. In the figure, the horizontal axis represents frequency and the vertical axis represents power. By the processing of the training signal extraction circuit 44-1, the component of the transmission data sequence included in the digital signal sequence is removed. The signal from which the transmission data sequence component has been removed from the first digital signal sequence and the signal from which the transmission data sequence component has been removed from the second digital signal sequence are combined by the multiplexing circuit 46 and signal power It is input to the calculation circuit 47 and the specific frequency signal component removal circuit 48.

  The signal power calculation circuit 47 determines, for each frequency component of the frequency spectrum input from the multiplexing circuit 46, whether the power of the frequency component is K times or more of the average power. The signal power calculation circuit 47 calculates the average value of the power of the frequency component having power K times or more of the average power. The signal power calculation circuit 47 outputs the calculated average value of the power as signal power to the SN ratio calculation circuit 50.

  The noise power calculation circuit 49 determines, for each frequency component of the frequency spectrum input from the specific frequency signal component removal circuit 48, whether the power of the frequency component is K times or more of the average power. The noise power calculation circuit 49 calculates an average value of frequency component power having power less than K times the average power. The noise power calculation circuit 49 outputs the average value of the calculated power to the SN ratio calculation circuit 50 as noise power. The noise power calculation circuit 49 may calculate the average value of the frequency component power from the signal input from the specific frequency signal component removal circuit 48 without performing the above determination.

  The SN ratio calculation apparatus 4 in the present embodiment is specified even when optical transmission by polarization multiplexing is performed or when a frequency band for measurement of noise power cannot be secured by high-density frequency multiplexing by the configuration and processing described above. It is possible to stably calculate and measure the SN ratio from the digital signal sequence including the frequency signal sequence components. In addition, since the specific frequency signal sequence (training signal sequence) used in this embodiment has low interference with the transmission data sequence, when performing optical transmission by polarization multiplexing, high density frequency multiplexing, Even when the transmission data series occupies a wide band due to high-speed modulation or the like, the SN ratio can be calculated stably.

  Further, in the optical transmission system according to the present embodiment, based on the SN ratio calculated by the SN ratio calculation device 4, the transmission format determining circuit 12-1 determines the modulation scheme, data rate, signal band, and FEC redundancy for the transmission data. And 12-2 are determined, it is possible to perform optical transmission according to the quality of the transmission path. That is, when the quality of the transmission path is high, the transmission capacity is increased, and when the quality of the transmission path is low, the transmission capacity is decreased to perform adaptive optical transmission according to the quality of the transmission path. it can. For example, in the case where QPSK is used as the modulation method, if the S / N ratio is higher than that capable of performing stable transmission, the modulation method is multi-valued to 8QAM or 16QAM, or the FEC redundancy is reduced. Thus, the frequency utilization efficiency is increased.

  Furthermore, when there are a plurality of optical transmission lines 2 in the optical transmission system, the control plane 5 transmits transmission data distributed to each optical transmission line 2 according to the SN ratio of each optical transmission line 2. May be. Further, based on the SN ratio calculated by the SN ratio calculation device 4 and the change in the data amount of the transmission data, it is possible to grasp and deal with a state in which there is no margin in the SN ratio in advance.

  In addition, when performing the FFT in the SN ratio calculation device 4, an overlap process is performed in which digital signals to be subjected to the FFT are selected by being overlapped. Thereby, the influence of the window edge which a signal distorts can be suppressed, and the precision at the time of estimating an S / N ratio can be improved. In addition, even when a window function that reduces window edge distortion is used, the window edge where the signal is distorted more than the window function can be removed, so that the SN ratio estimation accuracy can be improved.

  In the overlap FFT operation circuits 42-1 and 42-2 and the overlap FFT operation circuits 45-1 and 45-2 according to the present embodiment, the digital signal in the previous FFT is cut out in the extraction of the digital signal to be subjected to FFT. It is also possible to perform clipping using a window function such as a Hamming window or a Blackman window without performing an overlap process for selecting the signal to partially overlap.

In the present embodiment, the configuration using a specific frequency signal sequence in which power is concentrated on two frequencies has been described. However, there may be one frequency where power is concentrated in the specific frequency signal sequence, or four or more frequencies. However, when a specific frequency signal sequence in which power is concentrated on one frequency is used, a section in which the specific frequency signal sequence other than the frequency exists shifts when dispersion occurs. Therefore, a transmission data sequence is included when extracting noise. It may be lost. In this case, it is preferable to use a back-to-back in which no dispersion occurs or under extremely weak conditions, or transmission is not performed. For example, as a specific frequency signal series in which power is concentrated on four frequencies, (+ S, + S, -S, -S, + S, + S, -S, -S, ...) and one amplitude are repeated twice. The alternating signal is used.
In the present embodiment, the configuration in which two polarizations orthogonal to each other are multiplexed has been described.
However, optical transmission using a single polarization may be used.

  Further, in the present embodiment, the configuration for performing one-way optical transmission from the optical signal transmission device 1 to the optical signal reception device 3 has been described. However, an optical signal communication device including each circuit included in the optical signal transmission device 1 and each circuit included in the optical signal reception device 3 is provided at both ends of the optical transmission line 2 to perform bidirectional optical transmission. There may be. In this case, the S / N ratio calculation device 4 may be connected to any one of the optical signal communication devices to output the S / N ratio signal to the control plane 5, or the individual S / N ratio calculation device 4 may be provided to each optical signal communication device. May be connected, and each SN ratio calculation device 4 may output an SN ratio signal to the control plane 5.

  In the present embodiment, the configuration in which the SN ratio calculated by the SN ratio calculation device 4 is fed back to the optical signal transmission device 1, the optical signal reception device 3, and the control plane 5 has been described. However, the calculated SN ratio may be monitored to grasp the transmission state without feeding back the calculated SN ratio to the optical signal transmitting apparatus 1 and the optical signal receiving apparatus 3. In this case, the optical signal transmission device 1 may not include the SN ratio acquisition circuit 11 and the transmission format determination circuits 12-1 and 12-2. The modulation method, data rate, signal band, redundancy in forward error correction coding, and the like are determined in advance based on simulations and actual measurement values.

  In the present embodiment, the configuration in which the peak power time calculation circuit 43 specifies a section in which the specific frequency signal sequence is included in the digital signal sequence has been described. However, the peak power time calculation circuit 43 specifies the section of the three-dimensional data having the largest power of the frequency component corresponding to the specific frequency signal sequence based on the three-dimensional data, and sets the section as the first section. Also good. Accordingly, it is possible to suppress the amount of calculation in the processing subsequent to the peak power time calculation circuit 43.

  Further, in the present embodiment, when the signal power calculation circuit 47 and the noise power calculation circuit 49 determine whether or not they are components of a specific frequency signal sequence, the determination is based on the power that is K times the average power. The configuration has been described. However, the present invention is not limited to this, and it may be determined whether or not it is a component of a specific frequency signal sequence with reference to a predetermined threshold power. Further, when the signal power calculation circuit 47 calculates the power of the specific frequency signal sequence, the noise power calculated by the noise power calculation circuit 49 may be subtracted.

Further, in the present embodiment, each circuit of the SN ratio calculation device 4 has been described with respect to a configuration in which the frequency band of the transmission data sequence (for example, the frequency band from −25 GHz to 25 GHz in FIG. 2) is a processing target. The processing target in each circuit of the S / N ratio calculation device 4 is processed in a frequency band (for example, a frequency band from -12.5 GHz to 12.5 GHz in FIG. 3) including a frequency at which the power of the specific frequency signal sequence is concentrated. It may be a target. At this time, the noise power may be calculated using a frequency band near the intermediate frequency of the frequency at which the power of the specific frequency signal series is concentrated (for example, a 12.5 GHz band including 0 GHz in FIG. 3). As the band of 12.5 GHz, for example, a band from −6.75 GHz to 6.75 GHz is selected. This is to obtain an optical signal-to-noise ratio (OSNR), and OSNR is often used to represent the state of a signal in optical communication, and its definition is the noise power and signal within 12.5 GHz. It is the ratio with electric power. The OSNR is calculated using the following equation (1).
OSNR = (total signal power) / (noise power of 12.5 GHz band) (1)

  In the present embodiment, the signal power calculation circuit 47 and the noise power calculation circuit 49 have been described with respect to the configuration in which the average value of power is calculated in the processing target band to obtain the signal power and the noise power. However, the present invention is not limited to this, and the S / N ratio per sample may be calculated by obtaining the signal power and noise power per sample. At this time, the OSNR may be calculated by expanding or contracting to 12.5 GHz corresponding to the interval between samples. For example, when the interval between samples corresponds to 10 MHz, it is multiplied by 1250 and converted to a value corresponding to 12.5 GHz.

  In the present embodiment, the configuration in which the noise power calculation circuit 49 calculates the noise power based on the noise component in the section including the specific frequency signal sequence has been described. However, when the dispersion in the optical transmission line 2 is large, the speed of traveling in the optical transmission line 2 varies depending on the frequency. Therefore, there is a difference between a section in the digital signal series that includes the component of the specific frequency signal series and a section that includes the noise component to be used for calculating the SN ratio. In such a case, the section used for calculating the noise power may be determined based on the frequency at which the power of the specific frequency signal sequence is concentrated. Specifically, the section for calculating the noise power is determined as shown in FIG.

FIG. 12 is a diagram showing an outline of determination of a section used for calculation of noise power in the present embodiment. In the figure, the horizontal axis represents time and the vertical axis represents frequency. In the example shown in the figure, the power of the specific frequency signal series is concentrated on the frequencies fa and fb. The dispersion value at the frequency fa is Da, and the dispersion value at the frequency fb is Db. The section including only the specific frequency signal series varies depending on the frequency due to the influence of dispersion. The time ta at which the component of the specific frequency signal sequence appears at the frequency fa is different from the time tb at which the component of the specific frequency signal sequence appears at the frequency fb. However, by using the times ta and tb, it is possible to obtain the start time of the section in which the transmission data sequence is not included in the intermediate frequency between the frequency fa and the frequency fb as ((ta + tb) / 2). Further, dispersion at an intermediate frequency can also be obtained as ((Da + Db) / 2). By measuring the noise power in the section having the same length as the length of the specific frequency signal sequence with the time ((ta + tb) / 2) as the start time, the noise is not included in the transmission data sequence A or the transmission data sequence B. Electric power can be acquired. Further, not only the intermediate frequency ((fa + fb) / 2) but also using the times ta and tb and the variances Da and Db, noise power at frequencies other than the frequencies fa and fb can be acquired.
Further, by using the information on the times ta and tb, it is possible to compensate for the dispersion of the received optical signal. By performing dispersion compensation, the time difference of the optical signal for each frequency is eliminated, so that it is possible to extract a specific frequency signal sequence in the entire band without having to divide each frequency.

  In this case, the peak power time calculation circuit 43 specifies a section for each frequency at which the power of the specific frequency signal sequence is concentrated, and specifies a section for a frequency band for measuring noise power used when calculating the SN ratio. become. The training signal extraction circuits 44-1 and 44-2 individually extract sections corresponding to the respective frequencies of the specific frequency signal series and sections for noise, and output them to the overlap FFT calculation circuits 45-1 and 45-2. To do. The overlap FFT arithmetic circuits 45-1 and 45-2 perform processing separately for FFT for a specific frequency signal sequence and FFT for noise, and output the processing result to the multiplexing circuit 46. As a result, even when the variance is large, noise power can be obtained with high accuracy without including components of the transmission data sequence, and the accuracy of the SN ratio can be improved.

(Second Embodiment)
The SN ratio calculation device 4 according to the first embodiment is used when specifying a section of a specific frequency signal sequence included in the first digital signal sequence and the second digital signal sequence input from the optical signal reception device 3. The number of FFT points (window size) is smaller than the number of symbols of the specific frequency signal sequence. In the second embodiment, a configuration will be described in which the number of FFT points used when specifying a specific frequency signal sequence section is greater than the number of symbols in the specific frequency signal sequence. In the present embodiment, the number of frequencies where the power of the specific frequency signal sequence is concentrated is described as N. For example, N = 2 in the specific frequency signal sequence from which the frequency spectrum shown in FIG. 4 is obtained.

  FIG. 13 is a block diagram illustrating a configuration example of the SN ratio calculation apparatus 4A according to the second embodiment. The SN ratio calculation device 4A is provided in place of the SN ratio calculation device 4 in the optical transmission system (FIG. 1) described in the first embodiment. The signal to noise ratio calculation device 4A includes signal distribution circuits 401-1 and 401-2, overlap FFT circuits 402-1 and 402-2, signal distribution circuits 403-1 and 403-2, and frequency bandpass filter circuits 404-11 to 404-11. 404-1N and 404-21 to 404-2N, overlap IFFT circuits 405-11 to 405-1N and 405-21 to 405-2N, power calculation circuits 406-11 to 406-1N and 406-21 to 406-2N , Multiplexing circuits 407-1 to 407-N, averaging circuits 408-1 to 408-N, peak power time calculation circuit 409, training signal extraction circuits 44-1 and 44-2, overlap FFT operation circuit 45-1 45-2, multiplexing circuit 46, signal power calculation circuit 47, specific frequency signal component removal circuit 48, noise power calculation circuit 49 And, a SN ratio calculating circuit 50.

  Note that the training signal extraction circuits 44-1 and 44-2, the overlap FFT operation circuits 45-1 and 45-2, the multiplexing circuit 46, and the signal power, which perform processing subsequent to the peak power time calculation circuit 409, are performed. The calculation circuit 47, the specific frequency signal component removal circuit 48, the noise power calculation circuit 49, and the SN ratio calculation circuit 50 are the same as the circuits in the SN ratio calculation apparatus 4 (FIG. 6) of the first embodiment. Therefore, the same reference numerals are given and the description thereof is omitted. In the following, the signal distribution circuit 401-1, the overlap FFT operation circuit 402-1, the signal distribution circuit 403-1, the frequency band pass filter circuits 404-11 to 404-1N, and the overlap IFFT operation circuits 405-11 to 405-11. 405-1N and power calculation circuits 406-11 to 406-1N perform signal processing on the first digital signal series, and a signal distribution circuit 401-2, an overlap FFT operation circuit 402-2, and a signal distribution circuit 403- 2. Frequency band pass filter circuits 404-21 to 404-2N, overlap IFFT operation circuits 405-21 to 405-2N, and power calculation circuits 406-21 to 406-2N are used for signal processing on the second digital signal series. The structure which performs is demonstrated. In addition, each circuit that performs signal processing on the second digital signal sequence has the same configuration as each circuit that performs signal processing on the first digital signal sequence, so each circuit that performs signal processing on the second digital signal sequence The description of is omitted.

  The first digital signal series is input from the optical signal receiving device 3 to the signal distribution circuit 401-1. The signal distribution circuit 401-1 duplicates the input first digital signal sequence into two. The signal distribution circuit 401-1 outputs one of the two first digital signal sequences to the overlap FFT operation circuit 402-1, and trains the other first digital signal sequence. It outputs to the signal extraction circuit 44-1.

  Overlap FFT operation circuit 402-1 performs FFT on the first digital signal sequence output from signal distribution circuit 401-1, thereby converting a time domain signal to a frequency domain signal. At this time, the overlap FFT operation circuit 402-1 cuts out the first digital signal sequence so that the digital signal to be subjected to FFT overlaps with a part of the digital signal targeted in the previous FFT.

  As described above, the number of FFT points in the overlap FFT operation circuit 402-1 is set to be larger than the number of symbols of the specific frequency signal sequence. The overlap FFT operation circuit 402-1 performs FFT on the first digital signal sequence that is sequentially input, and the section or start time indicating the digital signal that is the subject of the FFT, and the frequency spectrum ( 3D data signal including frequency and power) is output to the signal distribution circuit 403-1.

  The signal distribution circuit 403-1 duplicates the three-dimensional data signal output from the overlap FFT operation circuit 402-1 to generate N three-dimensional data signals. The signal distribution circuit 403-1 outputs the generated N three-dimensional data signals to the frequency band pass filter circuits 404-11 to 404-1N, respectively.

  The frequency bandpass filter circuits 404-11 to 404-1N are bandpass filters that pass components in the vicinity of frequencies where the power of the specific frequency signal series is concentrated. The center frequency and the frequency width of the frequency band that each of the frequency band pass filter circuits 404-11 to 404-1N passes through are determined in advance based on a specific frequency signal sequence used in the optical transmission system. In addition, a different frequency is assigned to each circuit as the center frequency that each of the frequency band pass filter circuits 404-11 to 404-1N passes.

  For example, when the specific frequency signal sequence (N = 2) having the frequency spectrum shown in FIG. 4 is used, the SN ratio calculation device 4A uses the frequency bandpass filter circuits 404-11 and 404-12 for the first digital signal sequence. And frequency band pass filter circuits 404-21 and 404-22 for the second digital signal sequence. Further, the frequency band pass filter circuits 404-11 and 404-21 are configured as band pass filters that pass components in a frequency band having a central frequency of -12.5 GHz. The frequency band pass filter circuits 404-12 and 404-22 are configured as band pass filters that pass a frequency band component having a center frequency of 12.5 GHz.

  The frequency band pass filter circuit 404-11 has a frequency band centered on the frequency allocated to the own circuit with respect to the frequency spectrum included in the three-dimensional data signal output from the signal distribution circuit 403-1. Let the spectrum pass. The frequency band pass filter circuit 404-11 outputs information obtained by combining the passed frequency spectrum and the section or start time included in the three-dimensional data signal to the overlap IFFT arithmetic circuit 405-11. Further, the frequency band pass filter circuits 404-12 to 404-1N combine the passed frequency spectrum and the section or start time included in the three-dimensional data signal, similarly to the frequency band pass filter circuit 404-11. Information is output to overlap IFFT arithmetic circuits 405-12 to 405-1N

  The overlap IFFT arithmetic circuit 405-11 performs IFFT on the frequency spectrum included in the information output from the frequency bandpass filter circuit 404-11, and converts the frequency domain signal into a time domain signal. In the IFFT in the overlap IFFT arithmetic circuit 405-11, the frequency section targeted for IFFT is cut out from the frequency spectrum so that the frequency section targeted for the previous IFFT partially overlaps. That is, a process for overlapping IFFT targets is performed. The overlap IFFT arithmetic circuit 405-11 outputs information obtained by combining the digital signal sequence converted into the time domain signal and the section or start time included in the input information to the power calculation circuit 406-11.

  The overlap IFFT arithmetic circuits 405-12 to 405-1N, like the overlap IFFT arithmetic circuit 405-11, are frequency spectra included in information output from the frequency bandpass filter circuits 404-12 to 404-1N. Into a time domain signal. Overlapping IFFT arithmetic circuits 405-12 to 405-1N combine information obtained by combining a digital signal sequence converted into a time-domain signal and a section or start time included in input information into power calculation circuits 406-12 to 406-12. Output to 406-1N.

  The power calculation circuit 406-11 calculates a power value for each sample (symbol) with respect to the signal in the time domain included in the information output from the overlap IFFT arithmetic circuit 405-11. The power calculation circuit 406-11 outputs power information obtained by combining the calculated power value for each sample and the section or start time included in the input information to the multiplexing circuit 407-1. Similarly to the power calculation circuit 406-11, the power calculation circuits 406-12 to 406-1N perform time domain signals included in the information output from the overlap IFFT arithmetic circuits 405-12 to 405-1N. Then, the power value for each sample is calculated. The power calculation circuits 406-12 to 406-1 N combine power information obtained by combining the calculated power value after the sample and the section or start time included in the input information with the multiplexing circuits 407-2 to 407 -N. Output to.

  The multiplexing circuits 407-1 to 407-N are associated one-to-one with respect to the frequency at which the power of the specific frequency signal sequence is concentrated. The multiplexing circuits 407-1 to 407 -N are time-domain signals obtained from the first digital signal sequence and include a time-domain signal including a component in the vicinity of the frequency corresponding to the own circuit, and the second digital signal A time-domain signal obtained from the signal sequence and a time-domain signal including a component in the vicinity of the frequency corresponding to the own circuit is synthesized.

  The power information output from the power calculation circuit 406-11 and the power information output from the power calculation circuit 406-21 are input to the multiplexing circuit 407-1. The power information output from the power calculation circuit 406-21 is power information corresponding to the second digital signal series. The multiplexing circuit 407-1 performs synthesis by aligning the times based on the sections or start times included in each of the two pieces of power information and adding the power values for each sample included in each piece of power information. The multiplexing circuit 407-1 outputs the combined power information obtained by combining the power value for each sample obtained by combining and the section or the start time to the averaging circuit 408-1.

  Similarly to the multiplexing circuit 407-1, the multiplexing circuits 407-2 to 407-N include the power information output from the power calculation circuits 406-12 to 407-1N and the power calculation circuits 406-22 to 407-22. Power information output from 406-2N is input. Similarly to the multiplexing circuit 407-1, the multiplexing circuits 407-2 to 407-N align the time based on the sections or start times included in each of the two pieces of power information, and for each sample included in each piece of power information. The combination of adding the power values is performed. Similarly to the multiplexing circuit 407-1, the multiplexing circuits 407-2 to 407-N are circuits that average the combined power information obtained by combining the power value for each sample obtained by combining and the section or the start time. 408-2 to 408-N.

  As with the multiplexing circuits 407-1 to 407-N, the averaging circuits 408-1 to 408-N are associated one-to-one with the frequencies where the power of the specific frequency signal series is concentrated. The averaging circuit 408-1 calculates a moving average value for the power value for each sample included in the combined power information output from the multiplexing circuit 407-1. The averaging circuit 408-1 outputs average power information obtained by combining the calculated moving average value and the section or start time included in the combined power information to the peak power time calculation circuit 409. The averaging circuits 408-2 to 408 -N are similar to the averaging circuit 408-1 for the power value for each sample included in the combined power information output from the multiplexing circuits 407-2 to 407 -N. To calculate the moving average. The averaging circuits 408-2 to 408 -N, like the averaging circuit 408-1, use the average power information obtained by combining the calculated moving average value and the section or start time included in the combined power information as the peak power. The result is output to the time calculation circuit 409.

The average power information for each frequency at which the power of the specific frequency signal sequence is concentrated is input to the peak power time calculation circuit 409 from the averaging circuits 408-1 to 408 -N. The peak power time calculation circuit 409 specifies a section where the specific frequency signal sequence is arranged in the first digital signal sequence and the second digital signal sequence based on the moving average value of the power included in each average power information. To do. The peak power time calculation circuit 409 outputs section information indicating the identified section to the training signal extraction circuits 44-1 and 44-2.
Since the configuration of each circuit after the training signal extraction circuits 44-1 and 44-2 is the same as that of the first embodiment, the description thereof is omitted.

Here, the signal processing in the SN ratio calculation apparatus 4A will be further described.
Similarly to the SN ratio calculation apparatus 4 in the first embodiment, the SN ratio calculation apparatus 4A includes the first digital signal sequence having the frequency spectrum as shown in FIG. A digital signal sequence is input. A frequency spectrum as shown in FIG. 14 is obtained by signal processing by the overlap FFT operation circuit 402-1 and the frequency bandpass filter circuits 404-11 to 404-1N for the first digital signal sequence. FIG. 14 is a diagram illustrating an example of a frequency spectrum output from the frequency bandpass filter circuit 404-11 in the present embodiment. In the figure, the horizontal axis represents frequency and the vertical axis represents power. Further, the frequency spectrum shown in the figure is a frequency spectrum corresponding to a case where a digital signal sequence having the frequency spectrum shown in FIG. 8 is input to the SN ratio calculation device 4A.

  In the overlap IFFT arithmetic circuits 405-11 to 405-1N and 405-21 to 405-2N, the frequency spectrum as shown in FIG. 14 is converted into a time domain signal, and the power calculation circuits 406-11 to 406-1N and Signal processing in the time domain as shown in FIG. 15 is obtained by performing signal processing from 406-21 to 406-2N to the multiplexing circuits 407-1 to 407-N. FIG. 15 is a diagram illustrating an example of a time-domain signal input to the averaging circuits 408-1 to 408 -N in the present embodiment. In the figure, the horizontal axis represents the time axis and the vertical axis represents the power. Although the components corresponding to the transmission data sequence are suppressed by the filtering in the frequency band pass filter circuits 404-11 to 404-1 and 404-21 to 404-2N, the components of the transmission data sequence remain as shown in FIG. doing. If the S / N ratio is calculated using a signal in the time domain as shown in FIG. 15, the S / N ratio with a large error is calculated because the noise contains the component of the transmission data sequence.

  In the SN ratio calculation device 4A, the averaging circuits 408-1 to 408-N calculate the moving average value of the power value with respect to the signal in the time domain as shown in FIG. FIG. 16 is a diagram illustrating an example of a moving average value of power output from the averaging circuits 408-1 to 408 -N in the present embodiment. In the figure, the horizontal axis represents the time axis and the vertical axis represents the power. By obtaining the moving average value of the power value by the averaging circuits 408-1 to 408 -N, a triangular power value is obtained in the section where the specific frequency signal sequence is arranged. The peak power time calculation circuit 409 detects the peak power time at which the peak of the triangular power value that appears in the section where the specific frequency signal sequence is arranged. The peak power time calculation circuit 409 specifies a section corresponding to the number of symbols of the specific frequency signal sequence centered on the peak power time as the specific frequency signal sequence is arranged, and sets the section information indicating the section as a training signal extraction circuit. Output to 44-1 and 44-2.

  Based on the section information output from the peak power time calculation circuit 409, the training signal extraction circuits 44-1 and 44-2 have components of the specific frequency signal series in the first digital signal series and the second digital signal series. Extract the included digital signal sequence. As a result, the digital signal sequence extracted by the training signal extraction circuits 44-1 and 44-2 is a digital signal sequence not including the power of the transmission data sequence as shown in FIG. It becomes a digital signal sequence including only noise. Each circuit performs signal processing on this digital signal series, whereby the SN ratio can be calculated in the same manner as in the first embodiment.

  As described in the present embodiment, even if the number of digital signals (window size) to be subjected to FFT is larger than the number of symbols of the specific frequency signal sequence (training signal sequence), the first Similar to the embodiment, the S / N ratio related to the transmission data series included in the optical signal received by the optical signal receiving device 3 can be calculated.

  Note that the optical transmission system described in the present embodiment may be modified or modified as described in the first embodiment.

  In the present embodiment, two different frequency specific signal sequences may be combined as the specific frequency signal sequence generated by each of the specific frequency signal sequence generation circuits 14-1 and 14-2. Specifically, before and after the first specific frequency signal sequence having power concentrated in N specific frequency components, the second specific in which power is concentrated at a frequency different from the frequency at which the power of the first specific frequency signal sequence is concentrated. The signal sequence in which the frequency signal sequence is arranged is set as a specific frequency signal sequence to be output to the signal multiplexing circuits 15-1 and 15-2. In this case, the first specific frequency signal sequence is used to specify the section in which the training signal extraction circuits 44-1 and 44-2 cut out the digital signal sequence from the first digital signal sequence and the second digital signal sequence.

  FIG. 17 is a diagram illustrating an example of a specific frequency signal sequence using the first specific frequency signal sequence and the second specific frequency signal sequence. Such a specific frequency signal sequence (training signal sequence) is periodically time-division multiplexed with respect to the transmission data sequence. In this case, when extracting the frequency spectrum of the frequency band where the power of the first specific frequency signal sequence is concentrated in the frequency band pass filter circuits 404-11 to 404-1N and 404-21 to 404-2N, the second specific frequency is extracted. It is possible to remove the components of the signal sequence. In addition, since the first specific frequency signal sequence and the transmission data sequence are separated by the number of symbols of the second specific frequency signal sequence, residual transmission data sequence components can be suppressed. Thereby, the components of the transmission data sequence in the time domain signals input to the averaging circuits 408-1 to 408 -N can be reduced, and the first specific frequency signal sequence is arranged in the peak power time calculation circuit 409. It is possible to improve the accuracy of identifying a section that is present. Thereby, the calculation accuracy of the SN ratio can be improved. In addition, when the requirement for the accuracy of specifying the section in which the first specific frequency signal sequence is arranged is low, the second specific frequency signal sequence is arranged only before or after the first specific frequency signal sequence. May be.

(Third embodiment)
FIG. 18 is a block diagram illustrating a configuration example of the SN ratio calculation apparatus 4B according to the third embodiment. The SN ratio calculation device 4B of this embodiment is provided in place of the SN ratio calculation device 4 (FIG. 6) in the optical transmission system (FIG. 1) described in the first embodiment. The SN ratio calculation device 4B includes a training signal section detection block 40 and a power ratio calculation block 60.

  Similarly to the SN ratio calculation device 4, the SN ratio calculation device 4B receives the first digital signal sequence from the analog / digital conversion circuit 33-1 of the optical signal reception device 3, and the analog / digital conversion circuit 33-2. To the second digital signal sequence. The training signal section detection block 40 specifies a section in which the specific frequency signal series (training signal series) is included in the input first and second digital signal series. The power ratio calculation block 60 calculates the SN ratio between the transmission data series and the noise included in the first and second digital signal series based on the section specified by the training signal section detection block 40.

  FIG. 19 is a block diagram illustrating a configuration example of the training signal section detection block 40 in the present embodiment. As shown in the figure, the training signal section detection block 40 includes signal distribution circuits 401-1 and 401-2, overlap FFT circuits 402-1 and 402-2, signal distribution circuits 403-1 and 403-2, and a frequency. Band pass filter circuits 404-11 to 404-1N and 404-21 to 404-2N, overlap IFFT circuits 405-11 to 405-1N and 405-21 to 405-2N, power calculation circuits 406-11 to 406-1N 406-21 to 406-2N, multiplexing circuits 407-1 to 407-N, averaging circuits 408-1 to 408-N, and a peak power time calculation circuit 409.

  The circuits included in the training signal section detection block 40 are the circuits from the signal distribution circuits 401-1 and 401-2 included in the SN ratio calculation device 4 </ b> A (FIG. 13) in the second embodiment to the peak power time calculation circuit 409. It has the same configuration. Corresponding circuits are denoted by the same reference numerals, and description of the circuits is omitted. In the training signal section detection block 40, each circuit from the signal distribution circuits 401-1 and 401-2 to the peak power time calculation circuit 409 operates, so that the digital signal series obtained by duplicating the first and second digital signal series. And section information indicating a section including the specific frequency signal sequence (training signal sequence) in the digital signal sequence is output to the power ratio calculation block 60.

  FIG. 20 is a block diagram illustrating a configuration example of the power ratio calculation block 60 in the present embodiment. As shown in the figure, the power ratio calculation block 60 includes execution frequency giving circuits 61-1 and 61-2, signal extraction circuits 62-1 and 62-2, overlap FFT operation circuits 63-1 and 63-2, A multiplexing circuit 64, a specific frequency signal component removal circuit 65, a low frequency band component removal circuit 66, an overlap IFFT calculation circuit 67, a signal / noise power calculation circuit 68, a buffer circuit 69, and an SN ratio calculation circuit 70 are provided. Yes.

  The execution number giving circuit 61-1 counts the number of times (m) that the signal extraction circuit 62-1 has extracted a signal from the first digital signal series. Further, the execution number giving circuit 61-1 outputs the count value (m) to the signal extraction circuit 62-1. Similar to the execution number giving circuit 61-1, the execution number giving circuit 61-2 counts the number of times (m) that the signal extraction circuit 62-2 has extracted a signal from the second digital signal series. In addition, the execution number giving circuit 61-2 outputs the count value (m) to the signal extraction circuit 62-2.

  The signal extraction circuit 62-1 receives the first digital signal series and section information output from the training signal section detection block 40, and the count value (m) output from the execution number giving circuit 61-1. The signal extraction circuit 62-1 extracts a signal sequence from the first digital signal sequence based on the section information and the count value (m), and outputs the extracted signal sequence to the overlap FFT operation circuit 63-1. . The signal sequence extracted by the signal extraction circuit 62-1 is either a specific frequency signal sequence or a transmission data sequence. The signal extraction circuit 62-1 adds a count value (m) to the extracted signal series and outputs the result. This count value is used in the subsequent processing to identify whether the signal sequence is based on a specific frequency signal sequence or a transmission data sequence.

  The signal extraction circuit 62-2 receives the first digital signal series and section information output from the training signal section detection block 40, and the count value (m) output from the execution number giving circuit 61-2. Similarly to the signal extraction circuit 62-1, the signal extraction circuit 62-2 extracts a signal sequence from the second digital signal sequence based on the section information and the count value (m), and exceeds the extracted signal sequence. The result is output to the wrap FFT operation circuit 63-2. When the signal extraction circuit 62-1 extracts a signal sequence from a section of the training signal sequence, the signal extraction circuit 62-2 extracts a signal sequence from the section of the training signal sequence. When the signal extraction circuit 62-1 extracts a signal sequence from the transmission data sequence interval, the signal extraction circuit 62-2 extracts the signal sequence from the transmission data sequence interval. Similarly to the signal extraction circuit 62-1, the signal extraction circuit 62-2 gives a count value (m) to the extracted signal series and outputs it.

  FIG. 21 is a diagram illustrating a spectrum example of a digital signal series input to the signal extraction circuits 62-1 and 62-2 at the time of non-transmission. In the figure, the horizontal axis represents frequency and the vertical axis represents power (electric power). As shown in the figure, the digital signal series input to the signal extraction circuits 62-1 and 62-2 at the time of non-transmission includes a signal component, a noise component, and a training signal component.

  FIG. 22 is a diagram illustrating a spectrum example of a training signal component extracted during non-transmission. In the figure, the horizontal axis represents frequency and the vertical axis represents power (electric power). As shown in the figure, the signal sequence obtained by extracting the section of the specific frequency signal sequence in the signal extraction circuit 62-1 or 62-2 during non-transmission includes a noise component and a training signal component.

  It is determined based on the count value (m) whether the signal extraction circuit 62-1 extracts a signal sequence having a predetermined length from the specific frequency signal sequence or the transmission data sequence. For example, when calculating the S / N ratio based on 200 power measurements, a transmission data sequence is extracted when the count value (m) is an even number, and a specific frequency signal sequence is extracted when the count value (m) is an odd number. To do. Based on the section information input from the training signal section detection block 40, the signal extraction circuit 62-1 includes a training signal sequence in any section in the input first digital signal series, and training is performed in any section. It is determined whether a signal sequence is not included, that is, whether a transmission data sequence is included.

  Alternatively, the transmission data sequence may be extracted when the count value (m) is 1 to 100, and the specific frequency signal sequence may be extracted when the count value (m) is 101 to 200. That is, the signal extraction circuit 62-1 selects a specific frequency signal sequence and a transmission data sequence based on the count value (m) and extracts a signal sequence.

  The training signal section detection block 40 inputs time information indicating the time at which power is concentrated on a specific frequency corresponding to the specific frequency signal series to the signal extraction circuits 62-1 and 62-2 instead of the section information. In this case, the signal extraction circuits 62-1 and 62-2 operate as follows. The signal extraction circuits 62-1 and 62-2 extract the signal sequence of the training signal sequence length interval from the first digital signal sequence starting from the time indicated by the time information, or the signal sequence starting from the other interval. Is extracted from the first digital signal sequence as a transmission data sequence. For example, when the signal sequence length (number of points) extracted when calculating the power is 512, the time when the power is concentrated (peak power) in order to extract the signal sequence so as not to include the component of the specific frequency signal sequence Extraction is started from a point 512 points or more before (time), or extraction is started 512 points or more after the peak power time.

  The signal length (number of samples) of the signal sequences extracted by the signal extraction circuits 62-1 and 62-2 may be shorter than the training signal sequence length. Specifically, a training signal sequence is acquired at intervals of N samples from the start point indicated by the section information, and samples are extracted by a sample length that is 2N samples shorter than the training signal sequence. This is because extraction of a signal sequence that may overlap a training signal sequence (specific frequency signal sequence) and a transmission data sequence due to the influence of group delay or residual dispersion, which is a delay between polarizations occurring during transmission, is performed. This is to avoid it.

  For example, when group delay occurs, the section information (or the start time of the training signal sequence) between the first optical signal and the second optical signal calculated by the peak power time calculation circuit 409 is different. . When a signal with a signal bandwidth of 50 GHz is input to a PMD (Polarization Mode Dispersion) 0.5 ps / km ^ (1/2) optical fiber and transmitted at 1000 km, the group delay becomes 15 ps. . In the case of a system with a modulation rate of 32 GBaud, since one sample time is 31.25 ps, a signal sequence having a sample length shorter than that of the training signal sequence is extracted at the beginning and end of the training signal under the above transmission conditions. Thus, the influence of residual dispersion can be prevented.

  In addition, in the presence of residual dispersion, the transmission data sequence leaks into the training signal sequence, so that only the training signal sequence component cannot be extracted. For example, in the case of a system with a modulation rate of 32 GBaud and a signal band of 64 GHz, a residual dispersion of 50 ps occurs in the signal band when the residual dispersion is 100 ps / nm. Since one sample time is 31.25 ps, it is possible to prevent the influence of residual dispersion by extracting a sample length of about 2 samples at the beginning and end of the training signal sequence under the condition of the residual dispersion, and shorter than the training signal sequence. it can.

Returning to FIG. 20, the description of the configuration of the power ratio calculation block 60 will be continued.
The overlap FFT operation circuit 63-1 performs an FFT on the signal series output from the signal extraction circuit 62-1, thereby converting the signal in the time domain to the signal in the frequency domain. At this time, similar to the overlap FFT operation circuit 402-1, the overlap FFT operation circuit 63-1 performs a signal so that the digital signal to be subjected to FFT overlaps with a part of the digital signal targeted in the previous FFT. Cut out from the series. The overlap FFT operation circuit 63-1 outputs the frequency domain signal obtained by the FFT to the multiplexing circuit 64.

  Similar to the overlap FFT operation circuit 63-1, the overlap FFT operation circuit 63-2 performs FFT on the signal series output from the signal extraction circuit 62-2 to convert the time domain signal into the frequency domain. Convert to a signal. The overlap FFT operation circuit 63-2 outputs the frequency domain signal obtained by the FFT to the multiplexing circuit 64.

  The multiplexing circuit 64 synthesizes the frequency domain signals input from the overlap FFT arithmetic circuits 63-1 and 63-2, and outputs the frequency domain signal obtained by the synthesis to the specific frequency signal component removal circuit 65. To do. When a signal sequence is extracted from the training signal sequence in the signal extraction circuits 62-1 and 62-2, the frequency domain signal synthesized in the multiplexing circuit 64 and the frequency domain signal output from the multiplexing circuit 64 are extracted. The signal does not include a component of the transmission data sequence, but includes a training signal sequence and a noise component. In addition, when a signal sequence is extracted from the transmission data sequence in the signal extraction circuits 62-1 and 62-2, the frequency domain signal synthesized in the multiplexing circuit 64 and the frequency output from the multiplexing circuit 64 The signal in the region does not include a training signal sequence component, but includes a transmission data sequence and a noise component.

  The specific frequency signal component removal circuit 65 removes the frequency component of the specific frequency signal series from the frequency domain signal input from the multiplexing circuit 64 and outputs the result to the low frequency band component removal circuit 66. The frequency domain signal output from the specific frequency signal component removal circuit 65 becomes a signal including a noise component when the signal extraction circuit 62-1 and 62-2 extract the signal sequence from the training signal sequence. On the other hand, when a signal sequence is extracted from the transmission data sequence in the signal extraction circuits 62-1 and 62-2, the signal includes a component of the transmission data sequence and a noise component.

  The signal processing in the specific frequency signal component removal circuit 65 is performed in the frequency domain. The frequency component removed by the specific frequency signal component removal circuit 65 is a frequency component at which the power of the specific frequency signal series is concentrated. Since the frequency at which the power of the specific frequency signal series is concentrated is known, the specific frequency signal component removal circuit 65 may be configured by a bandpass filter to remove the frequency component at which the power is concentrated. Alternatively, the specific frequency signal component removal circuit 65 may be configured by a low-pass filter to extract a component in a low frequency band that does not include a frequency component at which power is concentrated.

  The low frequency band component removal circuit 66 removes the frequency component near the direct current from the frequency domain signal input from the specific frequency signal component removal circuit 65 and outputs the frequency component to the overlap IFFT operation circuit 67. The frequency range of the component removed by the low frequency band component removal circuit 66 is a range from direct current (0 Hz) to a predetermined frequency. For example, the predetermined frequency is determined based on a frequency component included in the transmission data series.

  The reason why the frequency component near the direct current is removed here is to remove the direct current component generated in the transmission signal when the bias driving the modulator in the optical signal transmission apparatus 1 fluctuates in time. This is because when measuring the noise power, the power of the frequency component near the direct current may be calculated to be larger than the actual power, thereby degrading the accuracy of the SN ratio calculation.

  FIG. 23 is a diagram illustrating a spectrum of a specific frequency signal sequence after transmission measured in an experiment. In the figure, the horizontal axis represents frequency and the vertical axis represents power. As shown in the figure, it can be observed that DC component power appears at a frequency corresponding to the carrier frequency. An example of a modulator used in optical communication is a Mach-Zehnder type optical modulator. FIG. 24 is a diagram illustrating a configuration of a Mach-Zehnder optical modulator. As shown in the figure, the Mach-Zehnder type optical modulator includes two amplitude modulators and one phase modulator, and performs single sideband modulation.

  Single sideband modulation is also referred to as single sideband modulation, and is a method of modulating in half the band as compared with double sideband modulation by combining two amplitude modulation components having a phase difference of 90 °. In the figure, an amplitude modulation signal is output from each of two amplitude modulators. One of the two amplitude modulation signals is modulated by the phase modulator so that the phase is 90 ° different from that of the other amplitude modulation signal. In order to cancel the DC component, the biases in the two amplitude modulators need to be appropriately controlled. However, since the bias driving the modulator fluctuates due to a temperature change or the like, it is difficult to completely remove the DC component. Furthermore, since the lasers on the transmission side and the reception side have a line width, the frequency is not a direct current but a frequency band corresponding to the line width.

  As a method for removing low frequency components near the direct current, there is a method using a band stop filter or a band pass filter. When a band stop filter is used, a filter having a frequency characteristic for suppressing a low frequency component near DC is applied to the low frequency band component removal circuit 66 for a frequency domain signal input to the low frequency band component removal circuit 66. To do. When a bandpass filter is used, a filter having a frequency characteristic that does not include a specific frequency signal sequence component and a low-frequency component near a direct current with respect to a frequency-domain signal input to the low-frequency band component removal circuit 66. Apply. When the band pass filter is used, the specific frequency signal component removal circuit 65 and the low frequency band component removal circuit 66 can be configured by one circuit. Similar to the specific frequency signal component removal circuit 65, the low frequency band component removal circuit 66 performs signal processing in the frequency domain.

Returning to FIG. 20, the description of the configuration of the power ratio calculation block 60 will be continued.
The overlap IFFT arithmetic circuit 67 converts the frequency domain signal input from the low frequency band component removal circuit 66 into a time domain signal by performing IFFT. The overlap IFFT operation circuit 67 outputs a time-domain signal obtained by IFFT to the signal / noise power calculation circuit 68. In the IFFT in the overlap IFFT operation circuit 67, as in the overlap IFFT operation circuit 405-11, the frequency spectrum so that the frequency section targeted for IFFT overlaps with the frequency section targeted for the previous IFFT. Cut out from.

The signal / noise power calculation circuit 68 calculates the power of the time domain signal input from the overlap IFFT calculation circuit 67. When the input time domain signal is a signal based on a training signal sequence (specific frequency signal sequence), the signal / noise power calculation circuit 68 stores the calculated power in the buffer circuit 69 as noise power. In addition, when the input time domain signal is a signal based on the transmission data series, the signal / noise power calculation circuit 68 stores the calculated power in the buffer circuit 69 as signal power.
The buffer circuit 69 stores the signal power and noise power calculated by the signal / noise power calculation circuit 68.

  When the signal / noise power calculation circuit 68 calculates a predetermined number of times (M) of power, that is, when the signal power and noise power are calculated, the S / N ratio calculation circuit 70 is stored in the buffer circuit 69. The S / N ratio is calculated based on the signal power and the noise power. When calculating the SN ratio, the SN ratio calculation circuit 70 reads the signal power stored in the buffer circuit 69 to calculate the average value of the signal power, and reads the noise power stored in the buffer circuit 69. The average value of noise power is calculated. Further, since the signal power stored in the buffer circuit 69 includes the power of the noise component, the SN ratio calculation circuit 70 subtracts the average value of the noise power from the average value of the signal power, The SN ratio is calculated from the average value of the noise power. In addition, when the average value of noise power is smaller than the average value of signal power, the SN ratio may be calculated without performing the above-described subtraction.

  FIG. 25 is a flowchart showing an SN ratio calculation process performed by the SN ratio calculation apparatus 4B in the present embodiment. In the S / N ratio calculation device 4B, when the S / N ratio calculation process is started, the execution frequency assignment circuits 61-1 and 61-2 initialize the counter by substituting “1” for the count value m (step S1). S11).

  The training signal section detection block 40 detects a section including the training signal series in each of the first digital signal series and the second digital signal series input from the optical signal receiving device 3. The training signal section detection block 40 outputs section information indicating the detected section, and the first digital signal sequence and the second digital signal sequence to the power ratio calculation block 60 (step S12).

  In the power ratio calculation block 60, the signal extraction circuits 62-1 and 62-2, based on the count value (m) output from the execution number giving circuits 61-1 and 61-2, A training signal sequence or a transmission data sequence is extracted from the second digital signal sequence (step S13).

  The signal sequences extracted by the signal extraction circuits 62-1 and 62-2 are converted into frequency domain signals by the overlap FFT calculation circuits 63-1 and 63-2. These frequency domain signals are synthesized by the multiplexing circuit 64, the frequency components of the training signal series are removed by the specific frequency signal component removal circuit 65, and the components near the direct current are removed by the low frequency band component removal circuit 66. . The frequency domain signal from which the component near DC is removed is further converted into a time domain signal by the overlap IFFT arithmetic circuit 67, and the power is calculated by the signal / noise power calculation circuit 68 (step S14).

The power value calculated by the signal / noise power calculation circuit 68 is identified based on the assigned count value (m). The signal / noise power calculation circuit 68 stores the calculated power value in the buffer circuit 69 as either signal power or noise power (step S15).
The SN ratio calculation circuit 70 determines whether or not the number of power value calculations (m) has reached a predetermined number (M) (step S16). If the number of calculations has not reached the predetermined number (step S16). : NO), the count value (m) in the execution frequency assignment circuits 61-1 and 61-2 is incremented by "1" (step S17), the process returns to step S13, and the signal series extraction and the power value calculation are performed. Let it be repeated.

  On the other hand, when the number of executions reaches a predetermined number (step S16: YES), the SN ratio calculation circuit 70 calculates the average values of the signal power and the noise power stored in the buffer circuit 69 (step S18). Then, the SN ratio is calculated from the calculated average value (step S19), and the SN ratio calculation process is terminated.

  As described above, the SN ratio calculation apparatus 4B switches the target between the transmission data sequence and the training signal sequence when calculating the power value, thereby calculating the transmission power from the transmission data sequence, and the training signal. The circuit can be shared with a circuit for calculating noise power from the series, and the circuit scale can be reduced. In addition, since the S / N ratio calculation device 4B includes the low frequency band component removal circuit 66, it is possible to remove components near the direct current that are included in the optical signal in the optical signal transmission device 1, and to measure noise power. The accuracy can be stabilized and the SN ratio accuracy can be improved.

  Further, in the SN ratio calculation device 4B, an overlap IFFT calculation circuit 67 is provided in front of the signal / noise power calculation circuit 68, and calculation processing in the training signal section detection block 40, calculation processing in the power ratio calculation block 60, The procedure is the same. That is, the signal processing is performed in the order of overlap FFT processing, filter processing, overlap IFFT processing, and power calculation. Thereby, it is possible to share each circuit in the training signal section detection block 40 and the power ratio calculation block 60, and it is possible to reduce the circuit scale of the SN ratio calculation device 4B.

  FIG. 26 is a diagram illustrating the effect of reducing the circuit scale in the SN ratio calculation apparatus 4B of the present embodiment. Each block in the figure is given a symbol corresponding to each circuit in FIGS. 19 and 20. As shown in FIG. 19 and FIG. 20, when calculating the SN ratio by obtaining the signal power corresponding to the transmission data series and the noise power of the noise component, the circuit configuration and noise power for calculating the signal power are calculated. The circuit configuration for calculation overlapped. Specifically, the calculation processing from the overlap FFT calculation circuits 402-1 and 402-2 to the power calculation circuits 406-11 to 406-1N and 406-21 to 406-2N in the training signal section detection block 40, and the power The calculation processing from the overlap FFT calculation circuits 63-1 and 63-2 to the signal / noise power calculation circuit 68 in the ratio calculation block 60 overlapped.

  Similarly, in the upper diagram of FIG. 26, the circuit from the FFT calculation circuit 402 to the power calculation circuit 406 in the TS section detection unit overlaps the circuit from the FFT calculation circuit 63 to the power calculation circuit 68 in the signal power and noise power calculation circuit. doing. Since the FFT operation circuit 402 to the power deriving circuit 406 and the FFT operation circuit 63 to the power calculation circuit 68 in the power ratio calculation block 60 have different time series, it is easy to share the circuits. Therefore, in the SN ratio calculation device 4B of the present embodiment, the execution number giving circuits 61-1 and 61-2 count the number of power calculations, and the signal extraction circuits 62-1 and 62-2 correspond to the number of calculations. A training signal sequence and a transmission data sequence are switched and extracted.

  Thereby, the training signal section detection block 40 and the power ratio calculation block 60 shown in FIG. 18 can be shared, and the circuit scale can be reduced. Specifically, in the training signal section detection block 40, calculation processing from the overlap FFT calculation circuits 402-1 and 402-2 to the power calculation circuits 406-11 to 406-1N and 406-21 to 406-2N, In the power ratio calculation block 60, the calculation processing from the overlap FFT calculation circuits 63-1 and 63-2 to the signal / noise power calculation circuit 68 is made common, and further reduction is achieved by performing signal processing in time division. Can be planned.

  The lower diagram of FIG. 26 shows the configuration in which the FFT 402, BPF 404, IFFT 405, and power derivation 406 for TS section detection are shared with the FFT 63, BPF 65, 66, IFFT 67, and power derivation 68 for signal power, respectively. You may share things related to noise power, you may share things related to signal power and noise power, or you may share three things related to TS section detection and signal power and noise power. Good. In addition, since the signal power side can be extracted from inside and outside the TS section, it is described as TS outside extraction 62 in the upper diagram of FIG. 26, but the noise extraction side TS extraction 62 and the signal extraction circuit 62 in FIG. Good.

  In the present embodiment, dispersion compensation may be performed on the first digital signal sequence and the second digital signal sequence input to the SN ratio calculation device 4B. As a dispersion compensation method, there are a method of compensating on the optical transmission line 2 using a dispersion compensation fiber, a method of compensating in the digital domain, a method of performing predistortion in the optical signal transmission apparatus 1, and the like. The method may be used. When dispersion compensation is performed on the optical transmission line 2, a method other than a method using a dispersion compensating fiber may be used.

Further, in the SN ratio calculation device 4B of the present embodiment, the signal power calculation and the noise power calculation are performed based on the count value (m) output from the execution number giving circuits 61-1 and 61-2. Although described, it may be performed alternately every frame. In this case, when OTU4 of 32640 symbols / frame and modulation rate of 32 GBaud are used, calculation of signal power and calculation of noise power are switched every 1.02 × 10 ⁻⁶ seconds.

  Further, in the SN ratio calculation device 4B of the present embodiment, the configuration in which the signal power and the noise power are calculated alternately has been described. However, the signal power and the noise power are calculated from the time domain signal calculated by the overlap IFFT arithmetic circuit 67. May be calculated simultaneously.

  In the SN ratio calculation device 4B of the present embodiment, the configuration in which the signal power and noise power calculated by the signal / noise power calculation circuit 68 are stored in the buffer circuit 69 as they are has been described. You may make it memorize | store in the buffer circuit 69, and may make it memorize | store in the buffer circuit 69 the value obtained by the moving average and the exponential average.

  Further, in the S / N ratio calculation device 4B of the present embodiment, the measurement time and measurement frequency of the signal power and the noise power are made equal. However, when the noise power is small compared to the signal power, the noise power measurement time (number of samples) or the measurement frequency is increased according to the power ratio in order to make the measurement accuracy comparable between the signal power and the noise power. May be.

  Further, in the SN ratio calculation device 4B of the present embodiment, an overlap IFFT arithmetic circuit 67 is provided in the power ratio calculation block 60, and signal power and noise power are calculated from a digital signal sequence converted into a time signal by IFFT. . However, the power ratio calculation block 60 may be configured not to include the overlap IFFT arithmetic circuit 67 but to calculate the signal power and the noise power from the frequency domain signal.

  In addition, in the S / N ratio calculation apparatus 4B of the present embodiment, the configuration in which the calculation function for performing IFFT is shared in time division and the calculation function for calculating the power value is shared in time division in FIG. You may share only.

  In the SN ratio calculation device 4B of the present embodiment, the signal power is calculated from the digital signal sequence extracted by using the signal extraction units 62-1 and 62-2 from other than the section specified by the training signal section detection block 40. The configuration including the signal / noise power calculation circuit 68 and the S / N ratio calculation circuit 70 that calculates the ratio between the power of the transmission data sequence and the power of the noise from the signal power and the noise power is exemplified. The other configurations may be different, or at least one of the signal extraction units 62-1 and 62-2, the signal / noise power calculation circuit 68, and the SN ratio calculation circuit 70 may be the first or second embodiment. You may apply to.

  By using the S / N ratio calculation apparatus in each of the above embodiments, the signal-to-noise power ratio can be obtained regardless of the transmission by polarization multiplexing or the configuration in the transmission path in optical communication. Further, according to the SN ratio calculation apparatus 4 in the first embodiment, the overlap FFT calculation circuits 42-1 and 42-2 perform FFT with the number of points smaller than the number of symbols of the training signal sequence, so that the transmission data sequence is A training signal sequence can be extracted while preventing leakage. Therefore, it is possible to specify a section including the training signal sequence in the first digital signal sequence and the second digital signal sequence without providing a frequency band pass filter circuit, an overlap IFFT circuit, a power calculation circuit, or the like. it can.

  Further, according to the SN ratio calculation device 4A in the second embodiment, the number of overlap FFT points when specifying a section including the training signal sequence is larger than the number of symbols in the training signal sequence, so The number of lap FFT operations can be reduced. Further, according to the SN ratio calculation device 4B in the third embodiment, the number of overlap FFT points is larger than the number of training signal sequence symbols as in the second embodiment. The number of times can be reduced, and the circuit scale can be reduced.

  In the training signal extraction circuits 44-1 and 44-2 of the first and second embodiments, the first digital signal is the same as the signal extraction circuits 62-1 and 62-2 of the third embodiment. The signal sequence length extracted from the sequence and the second digital signal sequence may be a signal sequence shorter than the training signal sequence length. Thereby, the influence of dispersion | distribution residual can be suppressed. In the first and second embodiments, as in the third embodiment, a low frequency band component removal circuit 66 that removes components near the direct current may be provided. As a result, it is possible to remove a component in the vicinity of a direct current that is included in the optical signal in the optical signal transmission device 1, stabilize the noise power measurement accuracy, and improve the SN ratio accuracy.

In the first embodiment, the second embodiment, and the third embodiment, the specific frequency signal sequence described below is generated by the specific frequency signal sequence generation circuits 14-1 and 14-2. It may be. For example, a specific frequency signal sequence in which the envelope shape of the frequency spectrum of the transmission data sequence and the frequency spectrum of the specific frequency signal sequence are in a proportional relationship may be generated. FIG. 27 is a diagram illustrating an example of a specific frequency signal sequence in which the transmission data sequence and the frequency spectrum of the specific frequency signal sequence are in a proportional relationship. In the figure, the horizontal axis represents frequency and the vertical axis represents power. The figure shows a frequency spectrum of a specific frequency signal sequence (N = 8) in which power is concentrated on eight frequencies. By using such a specific frequency signal sequence, even in a situation where the degree of degradation of the SN ratio differs for each frequency depending on the characteristics of a filter or the like used in the optical transmission system, the SN that reflects the influence of different degradation at each frequency is reflected. The ratio can be calculated. Further, the accuracy of the calculated S / N ratio can be improved.
Note that the specific frequency signal sequence having the frequency spectrum as shown in FIG. 27 can be obtained by adjusting the amplitude value of the alternating signal. The envelope of the frequency spectrum of the transmission data series can be obtained by simulation.

  Further, a specific frequency signal sequence may be generated in which the density of the frequency interval at which the power of the specific frequency signal sequence concentrates is proportional to the power level of the transmission data sequence. FIG. 28 is a diagram illustrating an example of a frequency spectrum of a specific frequency signal sequence in which the density of intervals between frequencies where the power of the specific frequency signal sequence is concentrated and the power level of the transmission data sequence have a proportional relationship. In the figure, the horizontal axis represents frequency and the vertical axis represents power. The figure shows a frequency spectrum of a specific frequency signal sequence (N = 6) in which power is concentrated on six frequencies. Such a specific frequency signal sequence can adjust the density of frequency intervals where power is concentrated by changing the interval of alternating signals based on the frequency spectrum of the transmission data sequence signal.

  Depending on the filter used during transmission of the optical signal, the frequency spectrum of the specific frequency signal sequence may be narrowed together with the frequency spectrum of the transmission data sequence. In this case, if the signal power is estimated using a specific frequency signal sequence that has been narrowed and whose power level has changed, the accuracy is reduced. For example, in the case of using a specific frequency signal sequence having a spectrum in which power on each side is concentrated around a carrier wave, it is assumed that the power on the high frequency side is decreased by 3 dB due to the influence of the filter. If the power of one specific frequency signal sequence is 1, a total of 4 powers can be obtained when there is no filter effect, but only 3 powers can be obtained when there is a filter effect. As a result, the estimated signal power also becomes 3/4 of the power compared to when the filter is not affected. Although the power of the transmission data sequence is also reduced due to the influence of the filter, the power of the transmission data sequence is concentrated in the central part and often does not decrease to 3/4. In such a case, by using the specific frequency signal sequence as shown in FIG. 28, the influence of the frequency dependency of the power can be suppressed, and the accuracy of the calculated S / N ratio can be improved.

  Further, a specific frequency signal sequence in which the power level at each frequency at which the power of the specific frequency signal sequence is concentrated is proportional to the loss rate of the frequency characteristic in the optical transmission system may be generated. FIG. 29 is a diagram illustrating an example of a frequency spectrum of a specific frequency signal sequence in which the power level at each frequency at which the power of the specific frequency signal sequence is concentrated is inversely proportional to the loss rate of the frequency characteristic in the optical transmission system. In the figure, the horizontal axis represents frequency and the vertical axis represents power. The figure shows a frequency spectrum of a specific frequency signal sequence (N = 4) in which power is concentrated on four frequencies.

  The frequency characteristics of the optical transmission system are, for example, taps of FIR filters used in the optical signal receiving apparatus 3 in a computer simulation having the same frequency characteristics as the optical signal transmitting apparatus 1, the optical transmission path 2, and the optical signal receiving apparatus 3. It can be obtained from the coefficient. Alternatively, in the optical transmission system, a signal sequence having the same level of frequency components is transmitted in all frequency bands used for transmission of the transmission data sequence, and each frequency component is analyzed in the optical signal receiving device 3. May be. By using the specific frequency signal sequence as shown in FIG. 29, it is possible to equalize the power at each frequency at which the power of the specific frequency signal sequence is concentrated, and to suppress the influence of the frequency characteristics received when estimating the signal power. The accuracy of the calculated SN ratio can be improved.

  In the first embodiment, the overlap FFT units 42-1 and 42-2 perform FFT with the number of points smaller than the number of symbols of the specific frequency signal sequence, and the FFT with the number of points smaller than the number of symbols of the specific frequency signal sequence. The configuration in which the peak power time calculation circuit 43 specifies the section where the specific frequency signal sequence is arranged based on the signal in the frequency domain obtained in the above has been described. However, without being limited to this configuration, the overlap IFFT arithmetic circuits 42-1 and 42-2 may perform FFT with the number of points equal to or greater than the number of symbols of the specific frequency signal sequence. Further, the peak power time calculation circuit 43 may specify a section in which the specific frequency signal sequence is arranged based on a frequency domain signal obtained by FFT with the number of points equal to or greater than the number of symbols of the specific frequency signal sequence. Good.

  In the second embodiment, the overlap IFFT arithmetic circuits 405-11 to 405-1N and 405-21 to 405 are provided before the power calculation circuits 406-11 to 406-1N and 406-21 to 406-2N. The configuration for providing 2N has been described. However, without being limited to this configuration, the power calculation circuits 406-11 to 406-1N and 406-21 to 406 are not provided without providing the overlap IFFT arithmetic circuits 405-11 to 405-1N and 405-21 to 405-2N. -2N may calculate the power value from the signal in the frequency domain.

  In the third embodiment, the configuration in which the signal / noise power calculation circuit 68 calculates the signal power and the noise power based on the time domain signal converted by the overlap IFFT arithmetic circuit 67 has been described. However, the signal / noise power calculation circuit 68 is not limited to this configuration, and the signal / noise power calculation circuit 68 does not use the time domain signal converted by the overlap IFFT calculation circuit 67, and the frequency domain output circuit 66 outputs the frequency domain signal. Either or both of signal power and noise power may be calculated from the signal.

  In the third embodiment, as shown in FIG. 26, overlap IFFT arithmetic circuits 405-11 to 405-1N and 405-21 between the training signal section detection block 40 and the SN ratio calculation block 60 are used. 405-2N and the overlap IFFT calculation circuit 67 share the calculation function for performing IFFT, and the power calculation circuits 406-11 to 406-1N and 406-21 to 406-2N and the signal / noise power calculation circuit 68 Explained the configuration sharing the calculation function for calculating the power value. However, only one of the sharing of the calculation function for performing IFFT and the sharing of the calculation function for calculating the power value may be performed.

  In the third embodiment, based on the section information calculated in the training signal section detection block 40, the signal extraction circuits 62-1 and 62-2 specify the digital signal series in the section including the specific frequency signal series. A signal / noise power calculation circuit 68 calculates a noise power based on the former digital signal sequence and extracts a signal power based on the latter digital signal sequence. The structure to calculate was demonstrated. However, the present invention is not limited to this, and in the third embodiment, signal power and noise power may be calculated in the same manner as in the first embodiment and the second embodiment. Also in the first embodiment and the second embodiment, a similar configuration may be applied to extract two types of digital signal sequences, a section including a specific frequency signal sequence and a section not including it.

  A program for realizing the function of the SN ratio calculation apparatus in the above-described embodiment is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed. You may perform the process which calculates S / N ratio. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer system” includes a WWW system having a homepage providing environment (or display environment). The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included.

  The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, and what is called a difference file (difference program) may be sufficient.

DESCRIPTION OF SYMBOLS 1 ... Optical signal transmitter 2 ... Optical transmission line 3 ... Optical signal receiver 4, 4A, 4B ... SN ratio calculation device 5 ... Control plane 11 ... SN ratio acquisition circuit 12-1, 12-2 ... Transmission format decision circuit 13 -1, 13-2: Transmission data sequence generation circuit 14-1, 14-2: Specific frequency signal sequence generation circuit 15-1, 15-2 ... Signal multiplexing circuit 16-1, 16-2 ... Electric / optical conversion circuit DESCRIPTION OF SYMBOLS 17 ... Polarization multiplexing circuit 31 ... Polarization division circuit 32-1, 32-2 ... Optical / electrical conversion circuit 33-1, 33-2 ... Analog / digital conversion circuit 34 ... Adaptive equalization circuit 35-1, 35- 2 ... Demodulation circuit 41-1, 41-2 ... Signal distribution circuit 42-1, 42-2 ... Overlap FFT operation circuit 43 ... Peak power time calculation circuit 44-1, 44-2 ... Training signal extraction circuit 45-1 45-2 Oval FFT calculation circuit 46... Multiplexing circuit 47... Signal power calculation circuit 48... Specific frequency signal component removal circuit 49 .. noise power calculation circuit 50 .. SN ratio calculation circuit 40... Training signal section detection block 60. DESCRIPTION OF SYMBOLS 1, 61-2 ... Execution frequency giving circuit 62-1, 62-2 ... Signal extraction circuit 63-1, 63-2 ... Overlap FFT operation circuit 64 ... Multiplexing circuit 65 ... Specific frequency signal component removal circuit 66 ... Low Frequency band component removal circuit 67 ... Overlap IFFT calculation circuit 68 ... Signal / noise power calculation circuit 69 ... Buffer circuit 70 ... SN ratio calculation circuit 401-1 and 401-2 ... Signal distribution circuit 402-1 and 402-2 ... Over Lap FFT arithmetic circuit 403-1, 403-2 ... Signal distribution circuit 404-11, 404-12, 404-1N, 4 4-21, 404-2N: Frequency band pass filter circuits 405-11, 405-12, 405-1N, 405-21, 405-2N ... Overlapping IFFT arithmetic circuits 406-11, 406-12, 406-1N, 406-21, 406-22, 406-2N: power calculation circuit 407-1, 407-2, 407-N: multiplexing circuit 408-1, 408-2, 408-N: averaging circuit 409: peak power time Calculation circuit

Claims (14)

In an optical transmission system that performs transmission using an optical signal in which a specific frequency signal sequence in which power is concentrated at a specific frequency and a transmission data sequence are time-division multiplexed,
An FFT unit for converting a digital signal sequence obtained from the received optical signal into a frequency domain signal;
A section for identifying a section in which the specific frequency signal sequence is arranged in the digital signal sequence based on a power value of a frequency at which the power of the specific frequency signal sequence is concentrated in the frequency domain signal converted by the FFT unit A specific part,
A signal extraction unit for extracting a digital signal sequence in the section specified by the section specifying unit from the digital signal sequence;
A specific frequency signal component removing unit that removes and outputs the component of the specific frequency signal sequence from the digital signal sequence extracted by the signal extracting unit;
Based on the digital signal sequence output by the specific frequency signal component removal unit, a noise power calculation unit that calculates noise power that is the power of noise included in the received optical signal;
An optical transmission system comprising: a power ratio calculation unit that calculates a ratio of power of transmission data series and noise power based on the noise power calculated by the noise power calculation unit. The optical transmission system according to claim 1,
The FFT unit converts the digital signal sequence obtained from the received optical signal into a frequency domain signal by performing FFT with a number of points smaller than the number of symbols of the specific frequency signal sequence.
Alternatively, the section specifying unit specifies a section in the frequency domain signal in which power exceeding a threshold power is converted by the FFT section as a section in which the specific frequency signal sequence is arranged in the digital signal sequence. ,
Alternatively, the FFT unit converts a digital signal sequence obtained from the received optical signal into a frequency domain signal by performing FFT with a number of points smaller than the number of symbols of the specific frequency signal sequence, and the interval specifying unit Specifies a section where power exceeding a threshold power is present in a frequency domain signal converted by the FFT unit as a section where the specific frequency signal sequence is arranged in the digital signal sequence. Transmission system. In the optical transmission system according to claim 1 or 2,
A filter unit that extracts a signal in the vicinity of a frequency at which power of the specific frequency signal sequence is concentrated from a signal in a frequency domain converted by the FFT unit;
A first IFFT unit for converting the signal extracted by the filter unit into a time domain signal;
A power calculator that calculates, for each symbol, the power value of the signal in the time domain converted by the first IFFT unit;
An averaging circuit that calculates a moving average value of the power value calculated by the power calculation unit;
The section specifying unit is
An optical transmission system, wherein a section in which the specific frequency signal sequence is arranged is specified in the digital signal sequence based on a time when the moving average value calculated by the averaging circuit becomes maximum. In the optical transmission system according to any one of claims 1 to 3,
A signal sequence obtained by connecting the first specific frequency signal sequence in which power is concentrated at the specific frequency and the second specific frequency signal sequence in which power is concentrated at a frequency different from the frequency at which the power of the first specific frequency signal sequence is concentrated. An optical transmission system, further comprising: a specific frequency signal sequence generation unit that generates a specific frequency signal sequence as the specific frequency signal sequence. In the optical transmission system according to any one of claims 1 to 4,
A second IFFT unit for converting the digital signal sequence used by the noise power calculation unit to calculate the noise power into a time domain signal;
The noise power calculator is
The optical transmission system, wherein the noise power is calculated based on a time-domain signal converted by the second IFFT unit. The optical transmission system according to claim 5,
The first IFFT unit and the second IFFT unit share an arithmetic function for performing IFFT in a time division manner, or the arithmetic function for calculating a power value in the noise power calculation unit is shared in a time division manner for the identification. The calculation of the power of the frequency signal sequence and the calculation of the noise power are performed. Alternatively, the arithmetic function for performing the IFFT in the first IFFT unit and the second IFFT unit is shared in a time division manner, and the noise power calculation unit. An optical transmission system characterized in that a calculation function for calculating a power value at the same time is shared in a time-sharing manner to calculate the power of the specific frequency signal sequence and the noise power. In the optical transmission system according to any one of claims 1 to 6,
The signal extraction unit extracts a digital signal sequence from other than the section specified by the section specifying unit;
Alternatively, the noise power calculation unit calculates signal power that is power of a digital signal sequence including the transmission data sequence based on a digital signal sequence other than the interval specified by the interval specifying unit, and the power ratio calculation unit Calculates the ratio of the power of the transmission data sequence and the power of noise from the signal power and the noise power,
Alternatively, the signal extraction unit extracts a digital signal sequence from other than the section specified by the section specifying unit, and the noise power calculation unit is other than the section specified by the section specifying unit extracted by the signal extracting unit. A signal power that is a power of a digital signal sequence including the transmission data sequence based on the digital signal sequence of the transmission data sequence, and the power ratio calculation unit calculates the power of the transmission data sequence from the signal power and the noise power. An optical transmission system characterized by calculating a ratio of noise power. The optical transmission system according to any one of claims 1 to 7,
A low frequency band component removing unit that removes and outputs a low frequency band component from the digital signal sequence output by the specific frequency signal component removing unit;
The noise power calculator is
The optical transmission system, wherein the noise power is calculated based on a digital signal sequence output from the low frequency band component removing unit. In the optical transmission system according to any one of claims 1 to 8,
According to the ratio calculated by the power ratio calculation unit, at least one of a modulation scheme, a data rate, a signal band, and a redundancy used for a signal sequence in which the specific frequency signal sequence and the transmission data sequence are time-division multiplexed. A transmission format determination unit for determining
An optical signal transmitter that encodes and modulates the transmission data sequence based on the determination of the transmission format determiner;
An optical transmission system, further comprising: a demodulation unit that performs demodulation and decoding corresponding to a modulation scheme, a data rate, a signal band, and redundancy used in the optical signal transmission unit. The optical transmission system according to any one of claims 1 to 9,
A polarization multiplexing unit that multiplexes and outputs two optical signals corresponding to two different signal sequences in which the specific frequency signal sequence and the transmission data sequence are time-division multiplexed in orthogonal polarization states;
A polarization separation unit that separates the received optical signal into a first optical signal and a second optical signal whose polarization planes are orthogonal to each other;
The FFT unit is
An optical transmission system, wherein a digital signal sequence obtained from the first optical signal and a digital signal sequence obtained from the second optical signal are converted into signals in the frequency domain. In the optical transmission system according to any one of claims 1 to 10,
A specific frequency signal sequence generation unit configured to generate, as the specific frequency signal sequence, a signal sequence in which an envelope shape of a frequency spectrum of the transmission data sequence and power concentrated on the specific frequency are in a proportional relationship; And optical transmission system. In the optical transmission system according to any one of claims 1 to 10,
A specific frequency signal sequence generation unit for generating, as the specific frequency signal sequence, a signal sequence in which a power level of a frequency spectrum of the transmission data sequence and a density of the specific frequency interval are proportional to each other; Optical transmission system. In the optical transmission system according to any one of claims 1 to 10,
An optical transmission system further comprising: a specific frequency signal sequence generation unit that generates a signal sequence in which a power value concentrated on the specific frequency is a power value corresponding to a frequency characteristic in the own system as the specific frequency signal sequence . A signal processing method in an optical transmission system that performs transmission using an optical signal in which a specific frequency signal sequence in which power is concentrated at a specific frequency and a transmission data sequence are time-division multiplexed,
An FFT step for converting a digital signal sequence obtained from the received optical signal into a frequency domain signal;
A section that identifies a section in which the specific frequency signal sequence is arranged in the digital signal sequence based on a power value of a frequency at which the power of the specific frequency signal sequence is concentrated in the frequency domain signal converted by the FFT step Specific steps,
A signal extraction step of extracting a digital signal sequence in the section identified in the section identification step from the digital signal sequence;
A specific frequency signal component removing step of removing and outputting the component of the specific frequency signal sequence from the digital signal sequence extracted in the signal extracting step;
Based on the digital signal sequence output in the specific frequency signal component removal step, a noise power calculation step for calculating a noise power that is a noise power included in the received optical signal;
A signal processing method comprising: a power ratio calculation step of calculating a ratio between the power of the transmission data sequence and the noise power based on the noise power calculated in the noise power calculation step.

Images