WO2021079647A1 - 係数更新量出力装置、係数更新装置、係数更新量出力方法及び記録媒体 - Google Patents
係数更新量出力装置、係数更新装置、係数更新量出力方法及び記録媒体 Download PDFInfo
- Publication number
- WO2021079647A1 WO2021079647A1 PCT/JP2020/034526 JP2020034526W WO2021079647A1 WO 2021079647 A1 WO2021079647 A1 WO 2021079647A1 JP 2020034526 W JP2020034526 W JP 2020034526W WO 2021079647 A1 WO2021079647 A1 WO 2021079647A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- coefficient update
- update amount
- filter
- sample value
- filter coefficient
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 60
- 238000012545 processing Methods 0.000 claims abstract description 50
- 238000004891 communication Methods 0.000 claims abstract description 31
- 230000006978 adaptation Effects 0.000 description 56
- 238000009795 derivation Methods 0.000 description 34
- 238000010586 diagram Methods 0.000 description 32
- 230000003287 optical effect Effects 0.000 description 32
- 230000003044 adaptive effect Effects 0.000 description 30
- 230000010287 polarization Effects 0.000 description 20
- 230000005540 biological transmission Effects 0.000 description 18
- 230000000903 blocking effect Effects 0.000 description 16
- 238000007792 addition Methods 0.000 description 14
- 238000001914 filtration Methods 0.000 description 9
- 239000006185 dispersion Substances 0.000 description 8
- 230000001427 coherent effect Effects 0.000 description 7
- 238000000926 separation method Methods 0.000 description 7
- 238000012549 training Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012937 correction Methods 0.000 description 3
- 238000003780 insertion Methods 0.000 description 3
- 230000037431 insertion Effects 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 239000013307 optical fiber Substances 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 238000011478 gradient descent method Methods 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 229910052691 Erbium Inorganic materials 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- UYAHIZSMUZPPFV-UHFFFAOYSA-N erbium Chemical compound [Er] UYAHIZSMUZPPFV-UHFFFAOYSA-N 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/50—Transmitters
- H04B10/516—Details of coding or modulation
- H04B10/54—Intensity modulation
- H04B10/541—Digital intensity or amplitude modulation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/60—Receivers
- H04B10/61—Coherent receivers
- H04B10/612—Coherent receivers for optical signals modulated with a format different from binary or higher-order PSK [X-PSK], e.g. QAM, DPSK, FSK, MSK, ASK
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
- H03H17/00—Networks using digital techniques
- H03H17/02—Frequency selective networks
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
- H03H17/00—Networks using digital techniques
- H03H17/02—Frequency selective networks
- H03H17/06—Non-recursive filters
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03H—IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
- H03H21/00—Adaptive networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/60—Receivers
- H04B10/61—Coherent receivers
- H04B10/614—Coherent receivers comprising one or more polarization beam splitters, e.g. polarization multiplexed [PolMux] X-PSK coherent receivers, polarization diversity heterodyne coherent receivers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/60—Receivers
- H04B10/61—Coherent receivers
- H04B10/616—Details of the electronic signal processing in coherent optical receivers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/60—Receivers
- H04B10/61—Coherent receivers
- H04B10/616—Details of the electronic signal processing in coherent optical receivers
- H04B10/6165—Estimation of the phase of the received optical signal, phase error estimation or phase error correction
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/32—Carrier systems characterised by combinations of two or more of the types covered by groups H04L27/02, H04L27/10, H04L27/18 or H04L27/26
- H04L27/34—Amplitude- and phase-modulated carrier systems, e.g. quadrature-amplitude modulated carrier systems
- H04L27/38—Demodulator circuits; Receiver circuits
Definitions
- the present invention relates to an equalization technique at the time of receiving optical communication.
- Non-Patent Document 1 has attracted attention as a technique for improving performance over general multi-level QAM modulation and achieving a communication speed close to the Shannon limit.
- PCS is an abbreviation for Probabilistic Constellation Shipping.
- the transmission line constitutes an additive Gaussian noise channel
- the signal points of the input signal achieving the Shannon limit have a continuous Gaussian distribution (see Non-Patent Document 3).
- a modulation method in which the signal points have a continuous Gaussian distribution is not practical. Therefore, PCS aims to realize a nearly continuous Gaussian distribution of signal points by changing the generation probability of each signal point based on multi-level QAM modulation.
- FIG. 1 is a diagram showing an example of a generation probability distribution of signal points of a PCS signal based on 64QAM.
- FIG. 1A is the same as a normal 64QAM, and shows a case where the generation probabilities of all signal points are constant.
- the probability of generating signal points is non-uniform, similar to a Gaussian distribution.
- the signal point generation probability is similar to the Gaussian distribution in this way, a communication speed close to the Shannon limit is achieved as compared with a normal QAM signal. Make it possible.
- PCS signal-to-noise ratio
- SNR signal-to-noise ratio
- FIG. 1 (a) which is the same as the normal 64QAM, has the highest signal point entropy and communication speed.
- FIGS. 1 (b) and 1 (c) are obtained, the communication speed becomes lower and the transmission at a lower SNR becomes possible.
- the adjustment of the communication speed by changing the shape of the distribution of the signal point generation probability has an advantage that it can be realized without major changes in the system configuration as compared with the method of changing the coding rate of the error correction code. ..
- a normal QAM signal can be basically used for digital signal processing for demodulation on the receiving side such as equalization, polarization separation, and carrier phase compensation.
- FIG. 2 is a conceptual diagram showing a configuration example of a general optical communication system using a PCS signal.
- the optical communication system 100 includes an optical transmitter 110, a transmission line 120, and an optical receiver 130.
- the optical transmitter 110 includes a coding unit 111, an LD 112, and an optical modulator 113.
- LD is an abbreviation for Laser diode.
- the coding unit 111 inputs the coded data obtained by encoding the input data to the optical modulator 113.
- the coded data is divided into, for example, four series and input to the optical modulator 113 in parallel.
- FIG. 2 the description of FIG. 2 is interrupted, and a configuration example of the coding unit 111 will be described.
- FIG. 3 is a conceptual diagram showing a configuration example of the coding unit 111 shown in FIG.
- the coding unit 111 Since the coding unit 111 generates a PCS signal based on the polarization multiplexing 64QAM, the orthogonal phase amplitude I of each of the first polarization (X polarization) and the second polarization (Y polarization) orthogonal to the first polarization (X polarization). And Q, a total of 4 sequences of signals are generated.
- I is an abbreviation for In-phase.
- Q is an abbreviation for Quadrature.
- the coding unit 111 includes an SP unit 21 and sub-coding units 22a to 22d.
- SP is an abbreviation for Serial / Parallel conversion.
- the sub-coding unit 22a includes an SP unit 23, a DM unit 24, an FEC coding unit 25, and a MAP unit 26.
- DM is an abbreviation for Distribution matcher.
- FEC is an abbreviation for Forward Error Correction.
- Each of the sub-encoding units 22b to 22d has the same configuration as the sub-encoding unit 22a.
- the SP unit 21 converts the input binary data into four series of parallel data. Each of the four series of data is input to each of the sub-encoding units 22a to 22d in parallel.
- the SP unit 23 inputs the input data to the DM unit 24 with the amplitude data used for the amplitude configuration and the positive / negative data amplitude data used for the configuration of the code thereof.
- the DM unit 24 performs DM on the input amplitude data.
- the DM unit 24 uses, for example, a Constant composition distribution matching algorithm for this DM.
- the FEC coding unit 25 performs FEC coding on the DM data input from the DM unit 24 and the positive / negative data input from the SP unit 23.
- FEC is an abbreviation for Forward Error Correction, as described above.
- the DM data and positive / negative data after FEC coding are input to the MAP unit 26.
- the MAP unit 26 performs mapping to QAM data using FEC-encoded DM data and positive / negative data, and generates PCS data.
- the generated PCS data is input to the light modulator 113 shown in FIG.
- the LD 112 shown in FIG. 2 inputs a laser beam, which is CW light, to the light modulator 113.
- CW is an abbreviation for Continuous wave.
- the light modulator 113 modulates the CW light input from the LD 112 with the coded data input from the coding unit 111.
- the modulated optical signal is transmitted to the optical receiver 130 via the transmission line 120.
- the transmission line 120 transmits the optical signal input from the optical transmitter 110 to the optical receiver 130.
- the transmission line 120 is an optical transmission line composed of, for example, an optical fiber or an EDFA.
- EDFA is an abbreviation for Erbium Doped Optical Fiber Amplifier.
- the optical receiver 130 includes an LD 131, a coherent receiver 132, an ADC 133, and a demodulation / decoding unit 134.
- ADC is an abbreviation for Analog-to-digital converter.
- the LD 131 inputs LD light to the coherent receiver 132 as a so-called local oscillator.
- the coherent receiver 132 is, for example, a polarization diversity type coherent receiver.
- the coherent receiver 132 detects the optical signal sent from the optical transmitter 110 via the transmission line 120 using the LD light input from the LD 131, and corresponds to the orthogonal phase amplitude of each polarized light.
- the four series of received signals are input to the ADC 133.
- the ADC 133 converts each of the input 4 series of analog signals into a digital 4 series of received sample values by sampling, and inputs them to the demodulation / decoding unit 134.
- the sampling rate of the sampling is, for example, twice the symbol rate.
- the demodulation / decoding unit 134 performs demodulation / decoding signal processing in the digital region for each of the input four series of received sample values.
- the demodulation / decoding unit 134 restores and outputs the data transmitted by the optical transmitter 110 by the signal processing.
- FIG. 4 is a conceptual diagram showing a configuration example of the demodulation / decoding unit 134 shown in FIG.
- the demodulation / decoding unit 134 includes a demodulation unit 31, sub-decoding units 32a to 32d, and a PS unit 35.
- PS is an abbreviation for Parallel / Serial conversion.
- the sub-decoding unit 32a includes an FEC decoding unit 33 and a DDM unit 34.
- DDM is an abbreviation for Distribution deraum.
- Each of the sub-decoding units 32b to 32d has the same configuration as the sub-decoding unit 32a.
- the demodulation unit 31 is input with the received sample values of the four series from the ADC 133 in FIG.
- the demodulation unit 31 demodulates each of these received sample values.
- the received sample values of the four series after demodulation are input to the sub-decoding units 32a to 32d.
- a specific example of the demodulation unit 31 is shown in FIG. 5, and the description thereof will be described later.
- the demodulation / decoding unit 134 is configured by a computer or a processor as a hardware configuration, and the processing performed by the demodulation unit 31 is typically executed by a program or information.
- the FEC decoding unit 33 performs FEC decoding on the input received sample value.
- the FEC decoding unit 33 inputs the received sample value after decoding to the DDM unit 34.
- the DDM unit 34 performs DDM on the input received sample value after FEC decoding.
- the received sample value after DDM is input to the PS unit 35.
- the PS unit 35 converts the parallel received sample values that have been decoded by each decoding unit into serial received data and outputs the data.
- the received data corresponds to the data input to the SP unit 21 of FIG.
- FIG. 5 is a conceptual diagram showing a configuration example of the demodulation unit 31 shown in FIG.
- the demodulation unit 31 includes a front-end compensation unit 81, a wavelength dispersion compensation unit 82, and an adaptive equalization unit 83.
- the front-end compensation unit 81 compensates for distortion caused by the reception front end such as skew for each of the four series of reception sample values input from the ADC 133.
- the front-end compensation unit 81 inputs each received sample value after the compensation to the wavelength dispersion compensation unit 82.
- the wavelength dispersion compensation unit 82 compensates for the wavelength dispersion accumulated in the transmission line for each polarization of each input received sample value.
- the wavelength dispersion compensation unit 82 inputs each received sample value after the compensation to the adaptive equalization unit 83.
- the adaptive equalization unit 83 performs equalization, polarization separation, and polarization mode dispersion compensation for each input received sample value by adaptation equalization.
- the adaptation equalization unit 83 outputs the received sample value after the adaptation equalization to the sub-decoding units 32a to 32d.
- FIG. 6 is a block diagram showing an example of adaptive equalization processing performed by the adaptive equalization unit 83 shown in FIG.
- each of the four series of received sample values x 1I , x 1Q , x 2I, and x 2Q is input in parallel from the wavelength dispersion compensating unit of FIG.
- the received sample value x 1I is the received sample value for the in-phase carrier wave for the first polarization (X polarization).
- the received sample value x 1Q is a received sample value for the quadrature carrier wave for the first polarization.
- the received sample value x 2I is a received sample value for an in-phase carrier wave for a second polarized wave (Y polarized wave) which is a polarized wave forming a right angle to the first polarized wave.
- the received sample value x 2Q is a received sample value for the quadrature carrier wave for the second polarization.
- the adaptation equalization unit 83 sets the received sample values x1I and x1Q as complex numbers ( x1I ) + i ( x1Q ) and the received sample values x2I and x2Q as complex numbers (x2I) + i (x) as complex numbering 10. 2Q ).
- Adaptive equalizer 83 for complex (x 1I) + i (x 1Q), the FIRF11a and 11b, for complex (x 2I) + i (x 2Q), a FIRF11c and 11d, respectively perform.
- FIRF is an abbreviation for finite impulse response filtering.
- each FIRF11a to 11d are adaptive equalizer 83 by the filter coefficient h ij (i, j is 1 or 2) indicating that the update is performed.
- the adaptation equalization unit 83 performs addition 15a which adds the output by FIRF11a and the output by FIRF11c to generate a complex number (y 1I ) + i (y 1Q).
- the adaptation equalization unit 83 performs addition 15b which adds the output by FIRF11a and the output by FIRF11c to generate a complex number (y 2I ) + i (y 2Q).
- the adaptive equalizer 83 as the carrier phase compensation 13a, the complex number (y 1I) + i (y 1Q), performs carrier phase compensation, generates a complex number (y '1I) + i ( y' 1Q).
- the carrier phase compensation is performed to compensate for the phase difference between the signal light and the local oscillator.
- the carrier phase compensation is performed by a digital phase locked loop (PLL) or the like. Carrier phase compensation is described, for example, in Non-Patent Documents 4 and 5.
- the adaptation equalization unit 83 also performs carrier phase compensation for the complex number (y 2I + i (y 2Q) as the carrier phase compensation 13b, and generates a complex number ( y'2I ) + i (y' 2Q).
- the adaptation equalization unit 83 sets the real and imaginary parts of the input complex numbers ( y'1I ) + i ( y'1Q ) and (y' 2I ) + i (y' 2Q ) as the real and imaginary part separation 14. To separate. Then, the adaptive equalizer 83, received sample value y '1I, y' 1Q, outputs the received sample values of 4 series of y '2I and y' 2Q.
- the adaptive equalization unit 83 performs equalization, polarization separation, and carrier phase compensation as described above.
- the output four received sample values are divided into quadrature amplitudes for each of the first and second polarizations for decoding performed in the subsequent stage.
- the adaptive equalization unit 83 derives a coefficient update amount ⁇ h ij for updating the filter coefficient h ij used for each FIRF as the coefficient update 12.
- the adaptation equalization unit 83 determines the complex number (x 1I ) + i (x 1Q ) and (x 2I ) + i (x 2Q ) and the complex number (y 1I ) + i (y 1Q ) and (y 2I ) + i (y). 2Q ) is used.
- Adaptive equalizer 83 to the outlet, further use of the phase difference [Phi i occurring for each input received sample value obtained by the carrier phase compensating 13a and 13b.
- the coefficient update amount ⁇ h ij is a value that increases or decreases the filter coefficient h ij at that time by updating.
- each of the subscripts i and j is 1 or 2.
- the adaptation equalization unit 83 updates the filter coefficient h ij at that time by increasing or decreasing the derived coefficient update amount ⁇ h ij.
- the update of the filter coefficient hij is adaptive.
- the Decision directed least mean squares (DDLMS) described in Non-Patent Document 2 is known. Assuming that each of the two series of received sample values input as complex numbers in each FIRF is x j , and each of the received sample values output from the FIRF as a real part and an imaginary part is y i , the relationship between them is as follows. become.
- the subscripts i and j are both 1 and 2.
- k is an integer representing discrete time and symbol timing.
- m is an integer representing discrete time.
- T represents transposition.
- the filter coefficient hij is updated so that the expected value of is smaller. This is done by stochastic gradient descent. That is, each filter coefficient h ij is set.
- the input received sample value has a phase offset or frequency offset, even if the tentative judgment is performed as it is, it will not work. Therefore, the received sample value whose carrier phase is compensated at the time of tentative determination is used.
- Equation 3 the second term on the right side of Equation 3 is derived as the coefficient update 12.
- the second term on the right side is the above-mentioned coefficient update amount ⁇ h ij .
- FIG. 7 is a block diagram showing an example of processing for deriving the coefficient update amount performed by the adaptation equalization unit 83.
- the coefficient update amount derivation process is a process for deriving the coefficient update amount ⁇ h ij described above.
- the received sample value x j shown in FIG. 9 is a received sample value input to the filter. Further, the received sample value y i is a sample value of the filter output. Further, the phase difference ⁇ i is a phase difference derived when the carrier phase compensation 13a or 13b of FIG. 6 is performed with respect to the received sample value y i.
- the adaptation equalization unit 83 makes a tentative determination 41 based on the received sample value y i and the phase difference ⁇ i.
- the tentative determination 41 is the same as the symbol determination of a normal QAM signal.
- the adaptation equalization unit 83 performs a subtraction 42 that subtracts the received sample value y i from the result of the provisional determination 41. Then, the adaptation equalization unit 83 performs multiplication 43 by multiplying the received sample value x j by the value after the subtraction 42.
- the value obtained by multiplication 43 is multiplied by ⁇ as a constant multiple 44, and is output as a coefficient update amount ⁇ h ij of the filter coefficient h ij.
- Adaptation equalization / polarization separation by this DDLMS is performed blindly without directly using the transmitted information. Therefore, when the adaptation equalization by DDLMS is performed on the PCS signal, a problem may occur depending on the conditions.
- FIG. 8 is a constellation diagram showing an example of failure in adaptation equalization by DDLMS.
- the constellation diagram is for a received symbol after carrier phase compensation of the first polarized wave (X polarized wave).
- the transmitted PCS signal has the probability distribution shown in FIG. 1 (c).
- the transmitted PCS signal has a lower entropy of the signal point and a remarkably high probability of generating a signal point near the center, as compared with a normal 64QAM that uniformly generates a signal point.
- the constellation diagram of FIG. 8 also requires a distribution with a high probability of signal points near the center, which is equal to this, for demodulation, but this is not the case, and the phase offset and amplitude are large as a whole.
- An object of the present invention is to provide a filter coefficient update amount output device or the like that can suppress erroneous convergence of filter coefficients without reducing the data communication speed.
- the filter coefficient update amount output device of the present invention calculates a coefficient update amount which is a value for updating the filter coefficient of the digital filter included in the equalizer that equalizes the received sample value of the coherently received received signal by digital data processing. It is an output filter coefficient update amount output device, and is derived from the difference between the provisional determination result and the processed received sample value of the processed received sample value which is the received sample value filtered by the digital filter. Minimize the magnitude of the difference between the first output unit that outputs the first coefficient update amount, the statistical information of the processed received sample value for a certain time width, and the set value of the statistical information. A second output unit that outputs the second coefficient update amount derived from the gradient related to the filter coefficient, and a third output unit that outputs the coefficient update amount derived from the first coefficient update amount and the second coefficient update amount. And.
- the filter coefficient update amount output device or the like of the present invention can suppress erroneous convergence of the filter coefficient without reducing the data communication speed.
- the first embodiment is an embodiment relating to a method for deriving a coefficient update amount that can suppress erroneous convergence of filter coefficients used in time domain filtering.
- the optical communication system of the present embodiment is the one of the present embodiment described below in the coefficient update amount derivation process performed by the adaptive equalization unit 83 of the general optical communication system shown in FIGS. 1 to 5. It was used.
- the coefficient update amount derivation process is a process for deriving the coefficient update amount used for updating the filter coefficient hij by the coefficient update 12 of FIG.
- the received sample value is divided into blocks in a predetermined time frame, and the filter coefficient hij is updated for each block. To do. Updating of the filter coefficients h ij by DDLMS was blocked, the
- Equation 6 the second term on the right side of Equation 6 is the above-mentioned coefficient update amount ⁇ h ij . Also,
- l is an integer (symbol timing) representing a discrete time.
- the filter coefficient update is
- ⁇ i status is the difference between the statistical information derived for the received symbol of the i-polarized light (i is 1 or 2) of this block and the desired value determined from the transmitted PCS signal.
- the statistical information may be such that PCS signals having different signal entropies can be distinguished.
- Statistical information is, for example, average intensity as a simple matter.
- the desired value determined from the transmitted PCS signal for example, the statistical information itself of the transmitted PCS signal and the value that takes into account the changes that occur in the transmission line are used.
- the desired value is shared in advance on the transmitting / receiving side before communication, or information is transmitted from the transmitting side via the communication path during communication.
- the communication path may be the transmission line 120 shown in FIG. 2 or a communication line other than the transmission line 120.
- FIG. 9 is a block diagram showing an example of the above-mentioned coefficient update amount derivation process.
- the received sample values x j and y i and the phase difference ⁇ i of the embodiment are the same as those described in the description of FIG. 7.
- Adaptive equalization unit 83 of FIG. 5 performs blocking 57a and 57b on the received sample value x j and y i.
- Each of the blocking 57a and 57b is a process of dividing each of the received sample values x j and y i, which are continuous received sample values, into blocks for each time width.
- the adaptation equalization unit 83 calculates the coefficient update amount by the DDLMS as the DDLMS 53 from the received sample values x j and y i after the blocking 57a and the phase difference ⁇ i.
- the adaptation equalization unit 83 performs a constant multiple 52a of ⁇ times the result obtained by DDLMS53.
- the adaptation equalization unit 83 performs statistical processing 56 for each block on the blocked reception sample value y i.
- Statistical processing 56 is, for example, calculation of the average intensity for the received sample value y i.
- the adaptation equalization unit 83 derives the difference between the result by the statistical processing 56 and the desired value.
- the desired value is a predetermined set value determined by the transmitted PCS signal or the like.
- the adaptation equalization unit 83 uses the received sample value x j where the blocking 57a is performed, the received sample value y i where the blocking 57b is performed, and the difference obtained by the difference derivation 55, and the gradient by the gradient derivation 54. Is derived.
- the adaptation equalization unit 83 performs a constant multiple 52b of ⁇ 'fold with respect to the result of the gradient derivation 54.
- the adaptation equalization unit 83 performs addition 51 by adding the value after the constant multiple 52a and the value after the constant multiple 52b, and outputs the value after the addition 51 as the coefficient update amount ⁇ h ij.
- FIG. 10 is a block diagram showing a coefficient update process using the average strength of blocks of received symbols as statistical information.
- the average intensity P of the transmitted PCS signal is used as the desired value shown in FIG.
- Equation 7 the coefficient update amount ⁇ h ij of the filter coefficient h ij when the average intensity of the block of the received symbol is used as statistical information is derived.
- the blocking 57a and 57b, the constant multiples 52a and 52b, and the difference derivation 55 shown in FIG. 11 are the same as those shown in FIG.
- the provisional determination 41 is the same as that shown in FIG.
- the subtraction 59 is a subtraction of the received sample value y i from the received sample value y i after the provisional determination 41.
- the addition 61 is an addition for the input value.
- Each of the multiplications 60 and 62 is a multiplication for the input value.
- the average intensity derivation 56a is a derivation of the average intensity for each block for the blocked received sample value y i.
- the complex conjugate 58 is a process for deriving the complex conjugate of Equation 10 from the received sample value x j.
- ⁇ 63 is a process for obtaining the sum in Equation 10.
- the probability density of the received symbol can also be used as statistical information. In that case, be able to distinguish PCS signals of different signal entropy more accurately, it is possible to reduce the possibility of erroneous convergence of the coefficient updating quantity Delta] h ij. In that case, the probability density of the PCS signal is used as the desired value.
- the probability density of the received symbols in order to approach the desired probability density probability density of the received symbol of the received sample values y i after FIRF is derived in a differentiable manner for the filter coefficients h ij. Then, the filter coefficient hij is updated by the stochastic gradient descent method so as to minimize the magnitude of the difference between the probability density of the received symbol and the desired probability density.
- the adaptive equalization unit 83 shown in FIG. 5 derives the probability density of the blocked received symbol, compares it with the desired probability density, and compares the difference.
- the probability density of the received symbol Qy i (y), when the desired probability density and Pr (y), these differences can be determined by the Kullback-Leibler divergence and L 2 norm, such as a number of ways.
- L 2 norm such as a number of ways.
- the probability density Qy i (y) of the received symbol is determined by the kernel density estimation method.
- FIG. 12 is a block diagram showing a coefficient update amount derivation process when the probability density of received symbols for each block is used as statistical information.
- the coefficient update amount derivation process is a process of deriving the coefficient update amount when the update processing unit 45 of FIG. 7 updates the filter coefficient hij .
- the blocking 57a and 57b, DDLMS53, gradient derivation 54, and constant multiples 52a and 52b in FIG. 12 are the same as those shown in FIG.
- the adaptation equalization unit 83 shown in FIG. 5 performs the probability density estimation 56b for estimating the probability density Qy i (y) of the received symbol for each block as the statistical processing 56 shown in FIG.
- the adaptation equalization unit 83 derives the difference between the probability density Qy i (y) of the received symbol for each block and the desired probability density Pr (y) as the difference derivation 55.
- the method for deriving the coefficient update amount of the FIRF filter coefficient of the present embodiment is performed by adding information considering the difference between the statistical information regarding the output from FIRF and the desired value to the derivation of the coefficient update amount by general DDLMS. It is said. Therefore, the coefficient update amount derivation method makes it possible to update the filter coefficient so that the statistical information does not easily deviate from the desired value. Therefore, the coefficient update amount derivation method can suppress erroneous convergence of the FIRF filter coefficient so that the difference is widened.
- the second embodiment is an embodiment relating to a method for deriving a coefficient update amount, which can suppress erroneous convergence of filter coefficients when filter processing in the frequency domain is used for adaptation equalization using a PCS signal.
- a frequency domain filter is used instead of the time domain filters FIRF 11a to 11d shown in FIG.
- FIG. 13 is a block diagram showing an example of the adaptation equalization processing of the present embodiment performed by the adaptation equalization unit 83 shown in FIG.
- the received sample values x 1I , x 1Q , x 2I and x 2Q and the complexization 10 are the same as those shown in FIG.
- the adaptive equalization unit 83 of FIG. 5 performs overlap / blocking 71 for each of the complex numbers (x 1I ) + i (x 1Q ) and (x 2I ) + i (x 2Q) generated by the complex number conversion 10. ..
- the overlap / block 71 is a block of received sample values for each predetermined time width by the overlap method. This process is performed because the received sample value is aperiodic, whereas a periodic signal is assumed in the FFT (Fast Fourier Transform) performed in the next stage.
- the adaptive equalization unit 83 performs FFT on the received sample values (x 1I ) + i (x 1Q ) and (x 2I ) + i (x 2Q ) after blocking as FFT 72a and 72b.
- the adaptation equalization unit 83 performs a 2x2 filter processing of FDF73a to 73d on the received sample value after the FFT.
- FDF is an abbreviation for Frequency Digital Filter, which is a filter in the frequency domain.
- the adaptation equalization unit 83 adds the output by the FDF73a and the output by the FDF73c as the addition 74a. Further, the adaptive equalization unit 83 adds the output by the FDF73b and the output by the FDF73d as the addition 74b.
- the adaptation equalization unit 83 converts the received sample values after the additions 74a and 74b into the received sample values in the time domain by the IFFT 77a and 77b.
- IFFT is an abbreviation for Inverse Fast Fourier Transform.
- the adaptation equalization unit 83 removes the overlap performed by the overlap / blocking 71 for each received sample value after the IFFT as the overlap removal / serialization 78.
- the adaptive equalization unit 83 performs carrier phase compensation 13a and 13b for each received sample value after removing the overlap.
- the carrier phase compensations 13a and 13b are the same as in the case of FIG.
- the adaptation equalization unit 83 separates the real part and the imaginary part of each received sample value after carrier phase compensation as the real imaginary part separation 14, and receives four series of received sample values y'1I , y'1Q , y. ' 2I and y'2Q is output.
- the adaptation equalization unit 83 updates the filter coefficient used for each FDF as the coefficient update 76 in parallel with the above processing. Adaptive equalization unit 83, the update is performed and the received sample value before filtering, and each received sample values of the filtered, the phase difference [Phi i obtained during carrier phase compensation 13a and 13b.
- the update of the filter coefficient is performed by the adaptive equalization algorithm as in the case of the time domain filter (FIRF) of the first embodiment.
- FIRF time domain filter
- Adaptation equalization unit 83 performs Constraint 75 on the coefficient update amount. Constraint is a process of converting the filter coefficient used for updating into the time domain, replacing the portion where the impulse response is applied to the overlapping portion with 0, and returning it to the frequency domain. Constraint is performed to avoid the wraparound distortion due to the assumption of the periodic signal in the FFT if the time width of the impulse response in the time domain of the frequency domain filter is not within the overlap amount. It is a thing.
- FIG. 14 is a block diagram showing a filter coefficient update amount derivation process when the average intensity of blocks of received symbols is used as statistical information in the coefficient update 76 shown in FIG.
- the adaptive equalization unit 83 determines the received sample value in the frequency domain before the filtering process and the received sample value in the frequency domain after the filtering process. From, the coefficient update amount is derived. In this case, since the received sample value has already been blocked by the conversion to the frequency domain by the FFT, the received sample values x j and y i are blocked again as shown in FIG. There is no need to do.
- the adaptation equalization unit 83 makes a tentative determination 41 with respect to the received sample value y i after the IFFT based on the phase difference ⁇ i. Then, the adaptation equalization unit 83 subtracts the received sample value y i from the provisional determination result as the subtraction 59.
- the adaptation equalization unit 83 performs 0Pad processing for returning to the double oversample with respect to the received sample value after the subtraction 59 as 0Pad66. Then, the adaptation equalization unit 83 performs FFT on the received sample value after 0 Pad 66 as FFT 64a. The adaptive equalization unit 83 further multiplies the received sample value after the FFT by the constant ⁇ , with the constant multiple 52a.
- the adaptation equalization unit 83 derives the average intensity for each block for the received sample value yi after the Fourier transform as the average intensity derivation 56a. Then, the adaptation equalization unit 83 derives the difference between the average intensity of the received sample value y i after the IFFT and the average intensity P of the transmitted PCS signal as the difference derivation 55.
- the adaptive equalizer 83 as a multiplication 60 multiplies the difference obtained by the difference deriving 55 to receive the sample values y i after IFFT65. Then, the adaptation equalization unit 83 performs FFT on the received sample value after multiplication 60 as FFT64b. Then, the adaptive equalization unit 83 multiplies the received sample value after FFT64b by ⁇ 'with a constant multiple of 52b.
- the adaptation equalization unit 83 adds the received sample value after the constant multiple 52a and the received sample value after the constant multiple 52b as the addition 61.
- the adaptation equalization unit 83 multiplies the received sample value after the complex conjugate 58 by the received sample value after the addition 61 as the multiplication 62, and outputs it as a coefficient update amount.
- FIG. 15 is a diagram showing a constellation of experimental results when the filter coefficient is updated by using the coefficient update amount derivation method of the present embodiment.
- a condition in which filter misconvergence occurs is used in general adaptation equalization by DDLMS as in the constellation shown in FIG. 7, but a frequency domain filter is received as a filter and statistical information is received. It is about the case where the average intensity of the symbol is used.
- FIG. 15 is a constellation of the received symbol after carrier phase compensation of the first polarized wave (X polarized wave).
- the method for deriving the coefficient update amount of the present embodiment can suppress erroneous convergence of the filter coefficient when a filter in the frequency domain is used as the filter for adaptive equalization.
- the method for deriving the coefficient update amount does not require insertion of the training pattern into the transmission data at regular intervals. Therefore, the method for deriving the coefficient update amount can suppress erroneous convergence of the filter coefficient without reducing the data communication speed.
- FIG. 16 is a block diagram showing the configuration of the filter coefficient update amount output device 83x, which is the minimum configuration of the filter coefficient update amount output device of the embodiment.
- the filter coefficient update amount output device 83x is a filter coefficient update amount output device that outputs the coefficient update amount.
- the coefficient update amount is a value for updating the filter coefficient of the digital filter provided in the equalizer that equalizes the received sample value of the coherently received received signal by digital data processing.
- the filter coefficient update amount output device 83x includes a first output unit 83xa, a second output unit 83xb, and a third output unit 83xc.
- the first output unit 83xa minimizes the magnitude of the difference between the provisional determination result of the processed received sample value, which is the received sample value filtered by the digital filter, and the processed received sample value. As described above, the first coefficient update value derived from the gradient with respect to the filter coefficient is output.
- the second output unit 83xb is derived from the gradient with respect to the filter coefficient so as to minimize the magnitude of the difference between the statistical information of the processed received sample value for a certain time width and the set value for the statistical information.
- the second coefficient update amount is output.
- the third output unit 83xc outputs the coefficient update amount derived from the first coefficient update amount and the second coefficient update amount.
- the filter coefficient update amount output device 83x is derived from the gradient with respect to the filter coefficient so as to minimize the magnitude of the difference between the statistical information and the set value. Therefore, the filter coefficient update amount output device 83x can suppress the filter coefficient from being updated so that the received sample value after processing deviates from the set value.
- the filter coefficient update amount output device 83x does not require insertion of the training pattern into the transmission data at regular intervals. Therefore, the filter coefficient update amount output device 83x can suppress erroneous convergence of the filter coefficient without reducing the data communication speed.
- the filter coefficient update amount output device 83x exhibits the effects described in the section of [Effects of the Invention] according to the above configuration.
- a filter coefficient update amount output device that outputs a coefficient update amount, which is a value for updating the filter coefficient of a digital filter provided in an equalizer that equalizes the reception sample value of a coherently received received signal by digital data processing.
- the first output that outputs the first coefficient update amount derived from the difference between the provisional determination result of the processed received sample value, which is the received sample value filtered by the digital filter, and the processed received sample value.
- Second output section and A third output unit that outputs the coefficient update amount derived from the first coefficient update amount and the second coefficient update amount is provided.
- Filter coefficient update amount output device (Appendix 2) The filter coefficient update amount output device according to Appendix 1, wherein the set value is a desired value which is an expected value for the statistical information. (Appendix 3) The filter coefficient update amount output device according to Appendix 1 or Appendix 2, wherein the set value is based on a transmitted Probabilistic Constellation Shipping signal.
- (Appendix 4) The filter coefficient update amount output device according to any one of Supplementary note 1 to Supplementary note 3, wherein the statistical information is the average intensity of received symbols related to the blocked received sample value after processing.
- (Appendix 5) The filter coefficient update amount output device according to any one of Supplementary note 1 to Supplementary note 3, wherein the statistical information is a probability density derived by kernel density estimation of a blocked received symbol.
- (Appendix 6) The filter coefficient update amount output device according to any one of Supplementary note 1 to Supplementary note 5, wherein the digital filter performs filter processing in a time domain.
- (Appendix 7) The filter coefficient update amount output device according to any one of Supplementary note 1 to Supplementary note 5, wherein the digital filter performs filter processing in the frequency domain.
- (Appendix 8) A filter coefficient update device that updates the filter coefficient by the filter coefficient update amount output device described in any one of Supplementary note 1 to Supplementary note 7.
- (Appendix 9) An equalization device including the filter coefficient updating device described in Appendix 8 and the digital filter.
- (Appendix 10) A receiving device including the equalizing device described in Appendix 9.
- (Appendix 11) A communication system including the receiving device described in Appendix 10.
- the first output that outputs the first coefficient update amount derived from the difference between the provisional determination result of the processed received sample value, which is the received sample value filtered by the digital filter, and the processed received sample value. And Outputs the second coefficient update amount derived from the gradient with respect to the filter coefficient so as to minimize the magnitude of the difference between the statistical information of the processed received sample value for a certain time width and the set value for the statistical information. Do the second output and A third output is performed to output the coefficient update amount derived from the first coefficient update amount and the second coefficient update amount. Filter coefficient update amount output method. (Appendix 13) Coherently Equalizes the received sample value of the received signal by digital data processing.
- Filter coefficient update amount output program to be made The first output that outputs the first coefficient update amount derived from the difference between the provisional determination result of the processed received sample value, which is the received sample value filtered by the digital filter, and the processed received sample value. And the processing to do Outputs the second coefficient update amount derived from the gradient with respect to the filter coefficient so as to minimize the magnitude of the difference between the statistical information of the processed received sample value for a certain time width and the set value for the statistical information. Processing to perform the second output and A process of performing a third output that outputs the coefficient update amount derived from the first coefficient update amount and the second coefficient update amount, and Filter coefficient update amount output program that causes the computer to execute.
- the received sample value in Appendix 1 is, for example, the received sample value x j shown in FIGS. 9, 10, 11, 12 or 14, or the received sample value x 1I , x shown in FIG. 1Q , x2I or x2Q .
- the equalizing device is, for example, the adaptive equalizing unit 83 shown in FIG.
- the digital filter performs, for example, a filter process of any one of FIRF11a to 11d shown in FIG.
- the digital filter is, for example, one that performs a filtering process according to any one of FDF73a to 11d shown in FIG.
- the filter coefficient is, for example, the above-mentioned hij .
- the coefficient update amount is, for example, the above-mentioned ⁇ h ij .
- the filter coefficient update amount output device is, for example, a part of the adaptation equalization unit 83 shown in FIG. 5 that outputs the result of deriving the coefficient update amount of FIGS. 9, 10, 11 or 12. .
- the processed received sample value is, for example, the received sample value yi shown in FIGS. 9, 10, 11, 12 or 14, or the received sample value x1I , x shown in FIG. 1Q , x2I or x2Q .
- the provisional determination result is, for example, the result of the provisional determination 41 shown in FIG. 11 or FIG.
- the difference is, for example, a subtraction result by subtraction 59 shown in FIG. 11 or FIG.
- the first coefficient update amount is, for example, a constant multiple result by the constant multiple 52a shown in FIG. 11 or FIG.
- the first output unit is, for example, a part of the adaptive equalization unit 83 shown in FIG. 5 that performs the constant multiple 52a shown in FIG. 11 or 14.
- the certain time width is, for example, a time width blocked by the blocking 57a or 57b shown in FIGS. 9, 10, 11 or 12.
- the time width is, for example, the time width blocked by the overlap / blocking 71 shown in FIG.
- the statistical information is, for example, a processing result by the statistical processing 56 shown in FIG. 9 or FIG.
- the statistical information is, for example, a derivation result by the average intensity derivation 56a shown in FIG. 11 or FIG.
- the statistical information is, for example, an estimation result by the probability density estimation 56b shown in FIG.
- the set value is, for example, a desired value shown in FIG.
- the set value is, for example, the average intensity P in the time width of the received symbol according to the received sample value shown in FIG. 10, FIG. 11 or FIG.
- the set value is, for example, the desired probability density Pr (y) shown in FIG.
- the second coefficient update amount is, for example, a processing result by a constant multiple 52b shown in FIG. 11 or FIG.
- the second output unit is, for example, a part of the adaptation equalization unit 83 shown in FIG. 5 that outputs the processing result of the constant multiple 52b shown in FIG. 11 or FIG.
- the third output unit is, for example, a part of the adaptation equalization unit 83 shown in FIG. 5 which outputs the result of the addition 51 shown in FIG. 9, FIG. 10 or FIG.
- the desired value in Appendix 2 is, for example, the desired value shown in FIG. 9, the average intensity P in the time width of the received symbol related to the received sample value shown in FIG. 10, FIG. 11 or FIG.
- the Probabilistic Constellation Shipping signal in Appendix 3 is, for example, the PCS signal described above.
- the blocking in Appendix 4 is, for example, the blocking 57a or 57b shown in FIGS. 9, 10, 11 or 12, or the overlap / blocking 71 shown in FIG.
- the average intensity of the received symbol is, for example, the average intensity P of the received symbol related to the received sample value shown in FIG. 10, FIG. 11 or FIG. 14 in the time width.
- the probability density derived by the kernel density estimation in Appendix 5 is, for example, the result of deriving the average intensity derivation 56a shown in FIG. 11 or FIG.
- the one that performs the filtering process in the time domain in Appendix 6 is, for example, any one of FIRF11a to 11d shown in FIG.
- the one that performs the filtering process in the frequency domain in Appendix 7 is, for example, any one of FDF73a to 73d shown in FIG.
- the filter coefficient updating device in Appendix 8 updates the filter coefficient according to the coefficient updating amount output by the filter coefficient updating amount output device of any one of Supplementary notes 1 to 7, for example, as shown in FIG. It is a part of the adaptation equalization part to be done.
- the equalization device in Appendix 9 is an adaptation equalization unit 83 shown in FIG. 5 in which the filter coefficient is updated by the filter coefficient update device described in Appendix 8.
- the receiving device in Appendix 10 is the optical receiver 130 represented in Appendix 2 including the equalizing device in Appendix 9. Further, in the communication system in Appendix 11, for example, in the combination of the optical transmitter 110 and the optical receiver 130 shown in FIG. 2, the optical receiver 130 is used as the receiving device.
- filter coefficient update amount output method in Appendix 12 is, for example, the coefficient update amount output method shown in FIGS. 9, 10, 11, 12, or 14.
- the filter coefficient update amount output program in Appendix 13 is, for example, the above-mentioned program for causing the above-mentioned computer to output the coefficient update amount shown in FIGS. 9, 10, 11, 12, or 14. ..
- the present invention has been described above using the above-described embodiment as a model example. However, the present invention is not limited to the above-described embodiments. That is, the present invention can apply various aspects that can be understood by those skilled in the art within the scope of the present invention. This application claims priority on the basis of Japanese application Japanese Patent Application No. 2019-191623 filed on October 21, 2019, and incorporates all of its disclosures herein.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Electromagnetism (AREA)
- Computer Hardware Design (AREA)
- Mathematical Physics (AREA)
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Abstract
Description
第一実施形態は、時間領域のフィルタ処理に用いられるフィルタ係数の誤収束を抑え得る係数更新量導出方法に関する実施形態である。
[構成と動作]
本実施形態の光通信システムは、図1乃至図5に表される一般的な光通信システムの適応等化部83が行う係数更新量導出処理に、以下に説明される本実施形態のものを用いたものである。係数更新量導出処理は、図6の係数更新12によるフィルタ係数hijの更新に用いられる係数更新量を導出する処理である。
DDLMSによるフィルタ係数制御の予備収束として、Constant Modulus Algorithm等による制御が行われることがあるが、この基準振幅としてPCS信号の平均強度Pから求まる値を使用することもできる。
[効果]
本実施形態のFIRFのフィルタ係数の係数更新量導出方法は、一般的なDDLMSによる係数更新量の導出に、FIRFからの出力に関する統計情報と所望値との差を考慮した情報を加味して行われる。そのため、前記係数更新量導出方法は、フィルタ係数を、前記統計情報が所望値から乖離しにくいように更新することを可能にする。そのため、前記係数更新量導出方法は、前記差が広がるようにFIRFのフィルタ係数が誤収束することを抑え得る。
<第二実施形態>
第二実施形態は、PCS信号を用いた適応等化に周波数領域のフィルタ処理を用いた場合のフィルタ係数の誤収束を抑え得る、係数更新量導出方法に関する実施形態である。
[構成と動作]
本実施形態では、図6に表される時間領域のフィルタであるFIRF11a乃至11dの代わりに、周波数領域のフィルタが用いられる。図13は、図5に表される適応等化部83が行う本実施形態の適応等化処理の例を表すブロック図である。
[効果]
本実施形態の係数更新量の導出方法は、適応等化を行うためのフィルタに周波数領域のフィルタを用いた場合において、フィルタ係数の誤収束を抑え得る。
(付記1)
コヒーレント受信した受信信号の受信サンプル値をデジタルデータ処理により等化する等化装置が備えるデジタルフィルタのフィルタ係数を更新するための値である係数更新量を出力するフィルタ係数更新量出力装置であって、
前記デジタルフィルタによるフィルタ処理が行われた前記受信サンプル値である処理後受信サンプル値についての仮判定結果と前記処理後受信サンプル値との差から導出した第一係数更新量を出力する第一出力部と、
ある時間幅についての前記処理後受信サンプル値の統計情報と、前記統計情報についての設定値との差分の大きさを最小化するように前記フィルタ係数に関する勾配から導出した第二係数更新量を出力する第二出力部と、
前記第一係数更新量と前記第二係数更新量とから導出した前記係数更新量を出力する第三出力部とを備える、
フィルタ係数更新量出力装置。
(付記2)
前記設定値は、前記統計情報についての期待される値である所望値である、付記1に記載されたフィルタ係数更新量出力装置。
(付記3)
前記設定値は、送信されたProbabilistic Constellation Shaping信号に基づくものである付記1又は付記2に記載されたフィルタ係数更新量出力装置。
(付記4)
前記統計情報は、ブロック化された前記処理後受信サンプル値に係る受信シンボルの平均強度である、付記1乃至付記3のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
(付記5)
前記統計情報は、ブロック化した受信シンボルのカーネル密度推定により導出された確率密度である、付記1乃至付記3のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
(付記6)
前記デジタルフィルタは、時間領域のフィルタ処理を行うものである、付記1乃至付記5のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
(付記7)
前記デジタルフィルタは、周波数領域のフィルタ処理を行うものである、付記1乃至付記5のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
(付記8)
付記1乃至付記7のうちのいずれか一に記載されたフィルタ係数更新量出力装置により前記フィルタ係数を更新するフィルタ係数更新装置。
(付記9)
付記8に記載されたフィルタ係数更新装置と前記デジタルフィルタとを備える等化装置。
(付記10)
付記9に記載された等化装置を備える受信装置。
(付記11)
付記10に記載された受信装置を備える通信システム。
(付記12)
コヒーレント受信した受信信号の受信サンプル値をデジタルデータ処理により等化する等化装置が備えるデジタルフィルタのフィルタ係数を更新するための値である係数更新量を出力するフィルタ係数更新量出力方法であって、
前記デジタルフィルタによるフィルタ処理が行われた前記受信サンプル値である処理後受信サンプル値についての仮判定結果と前記処理後受信サンプル値との差から導出した第一係数更新量を出力する第一出力を行い、
ある時間幅についての前記処理後受信サンプル値の統計情報と、前記統計情報についての設定値との差分の大きさを最小化するように前記フィルタ係数に関する勾配から導出した第二係数更新量を出力する第二出力を行い、
前記第一係数更新量と前記第二係数更新量とから導出した前記係数更新量を出力する第三出力を行う、
フィルタ係数更新量出力方法。
(付記13)
コヒーレント受信した受信信号の受信サンプル値をデジタルデータ処理により等化する等化装置が備えるデジタルフィルタのフィルタ係数を更新するための値である係数更新量を出力するフィルタ係数更新量出力をコンピュータに実行させるフィルタ係数更新量出力プログラムであって、
前記デジタルフィルタによるフィルタ処理が行われた前記受信サンプル値である処理後受信サンプル値についての仮判定結果と前記処理後受信サンプル値との差から導出した第一係数更新量を出力する第一出力を行う処理と、
ある時間幅についての前記処理後受信サンプル値の統計情報と、前記統計情報についての設定値との差分の大きさを最小化するように前記フィルタ係数に関する勾配から導出した第二係数更新量を出力する第二出力を行う処理と、
前記第一係数更新量と前記第二係数更新量とから導出した前記係数更新量を出力する第三出力を行う処理と、
を前記コンピュータに実行させるフィルタ係数更新量出力プログラム。
以上、上述した実施形態を模範的な例として本発明を説明した。しかしながら、本発明は、上述した実施形態には限定されない。即ち、本発明は、本発明のスコープ内において、当業者が理解し得る様々な態様を適用することができる。
この出願は、2019年10月21日に出願された日本出願特願2019-191623を基礎とする優先権を主張し、その開示の全てをここに取り込む。
11a、11b、11c、11d FIRF
12 係数更新
13a、13b キャリア位相補償
14 実虚数部分離
15a、15b、51、61、74a、74b 加算
21、23 SP部
22a、22b、22c、22d 副符号化部
24 DM部
25 FEC符号化部
26 MAP部
31 復調部
32a、32b、32c、32d 副復号部
33 FEC復号部
34 DDM部
41 仮判定
43、60、62 乗算
44、52a、52b 定数倍
53 DDLMS
54 勾配導出
55 差分導出
56 統計処理
56a 平均強度導出
56b 確率密度推定
57a、57b ブロック化
58 x* j
59 減算
63 Σ
64a、64b FFT
65 IFFT
66 0Pad
71 オーバラップ/ブロック化
72a、72b FFT
73a、73b、73c、73d FDF
75 Constraint
76 係数更新
77a、77b IFFT
78 オーバラップ除去/シリアル化
81 フロントエンド補償部
82 波長分散補償部
83 適応等化部
83x フィルタ係数更新量出力装置
83xa 第一出力部
83xb 第二出力部
83xc 第三出力部
100 光通信システム
110 光送信機
111 符号化部
112、131 LD
113 光変調器
120 伝送路
130 光受信機
132 コヒーレント受信機
133 ADC
134 復調復号部
Claims (13)
- コヒーレント受信した受信信号の受信サンプル値をデジタルデータ処理により等化する等化装置が備えるデジタルフィルタのフィルタ係数を更新するための値である係数更新量を出力するフィルタ係数更新量出力装置であって、
前記デジタルフィルタによるフィルタ処理が行われた前記受信サンプル値である処理後受信サンプル値についての仮判定結果と前記処理後受信サンプル値との差から導出した第一係数更新量を出力する第一出力手段と、
ある時間幅についての前記処理後受信サンプル値の統計情報と、前記統計情報についての設定値との差分の大きさを最小化するように前記フィルタ係数に関する勾配から導出した第二係数更新量を出力する第二出力手段と、
前記第一係数更新量と前記第二係数更新量とから導出した前記係数更新量を出力する第三出力手段とを備える、
フィルタ係数更新量出力装置。 - 前記設定値は、前記統計情報についての期待される値である所望値である、請求項1に記載されたフィルタ係数更新量出力装置。
- 前記設定値は、送信されたProbabilistic Constellation Shaping信号に基づくものである請求項1又は請求項2に記載されたフィルタ係数更新量出力装置。
- 前記統計情報は、ブロック化された前記処理後受信サンプル値に係る受信シンボルの平均強度である、請求項1乃至請求項3のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
- 前記統計情報は、ブロック化した受信シンボルのカーネル密度推定により導出された確率密度である、請求項1乃至請求項3のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
- 前記デジタルフィルタは、時間領域のフィルタ処理を行うものである、請求項1乃至請求項5のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
- 前記デジタルフィルタは、周波数領域のフィルタ処理を行うものである、請求項1乃至請求項5のうちのいずれか一に記載されたフィルタ係数更新量出力装置。
- 請求項1乃至請求項7のうちのいずれか一に記載されたフィルタ係数更新量出力装置により前記フィルタ係数を更新するフィルタ係数更新装置。
- 請求項8に記載されたフィルタ係数更新装置と前記デジタルフィルタとを備える等化装置。
- 請求項9に記載された等化装置を備える受信装置。
- 請求項10に記載された受信装置を備える通信システム。
- コヒーレント受信した受信信号の受信サンプル値をデジタルデータ処理により等化する等化装置が備えるデジタルフィルタのフィルタ係数を更新するための値である係数更新量を出力するフィルタ係数更新量出力方法であって、
前記デジタルフィルタによるフィルタ処理が行われた前記受信サンプル値である処理後受信サンプル値についての仮判定結果と前記処理後受信サンプル値との差から導出した第一係数更新量を出力する第一出力を行い、
ある時間幅についての前記処理後受信サンプル値の統計情報と、前記統計情報についての設定値との差分の大きさを最小化するように前記フィルタ係数に関する勾配から導出した第二係数更新量を出力する第二出力を行い、
前記第一係数更新量と前記第二係数更新量とから導出した前記係数更新量を出力する第三出力を行う、
フィルタ係数更新量出力方法。 - コヒーレント受信した受信信号の受信サンプル値をデジタルデータ処理により等化する等化装置が備えるデジタルフィルタのフィルタ係数を更新するための値である係数更新量を出力するフィルタ係数更新量出力をコンピュータに実行させるフィルタ係数更新量出力プログラムであって、
前記デジタルフィルタによるフィルタ処理が行われた前記受信サンプル値である処理後受信サンプル値についての仮判定結果と前記処理後受信サンプル値との差から導出した第一係数更新量を出力する第一出力を行う処理と、
ある時間幅についての前記処理後受信サンプル値の統計情報と、前記統計情報についての設定値との差分の大きさを最小化するように前記フィルタ係数に関する勾配から導出した第二係数更新量を出力する第二出力を行う処理と、
前記第一係数更新量と前記第二係数更新量とから導出した前記係数更新量を出力する第三出力を行う処理と、
を前記コンピュータに実行させるフィルタ係数更新量出力プログラムが記録された記録媒体。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US17/766,296 US12009868B2 (en) | 2019-10-21 | 2020-09-11 | Coefficient update amount output device, coefficient update device, coefficient update amount output method, and recording medium |
JP2021554153A JP7327502B2 (ja) | 2019-10-21 | 2020-09-11 | フィルタ係数更新量出力装置、フィルタ係数更新装置、フィルタ係数更新量出力方法及びフィルタ係数更新量出力プログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2019-191623 | 2019-10-21 | ||
JP2019191623 | 2019-10-21 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2021079647A1 true WO2021079647A1 (ja) | 2021-04-29 |
Family
ID=75619799
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2020/034526 WO2021079647A1 (ja) | 2019-10-21 | 2020-09-11 | 係数更新量出力装置、係数更新装置、係数更新量出力方法及び記録媒体 |
Country Status (3)
Country | Link |
---|---|
US (1) | US12009868B2 (ja) |
JP (1) | JP7327502B2 (ja) |
WO (1) | WO2021079647A1 (ja) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP4002749B1 (en) * | 2020-11-20 | 2024-10-16 | Nokia Technologies Oy | Continuous variable quantum key distribution |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012118215A1 (ja) * | 2011-03-02 | 2012-09-07 | 日本電気株式会社 | 光受信器、偏波分離装置、および光受信方法 |
WO2018198891A1 (ja) * | 2017-04-25 | 2018-11-01 | 日本電気株式会社 | 信号合成装置及び信号合成方法 |
WO2019198465A1 (ja) * | 2018-04-12 | 2019-10-17 | 日本電気株式会社 | 光空間通信システム、光受信装置、光受信方法、及び非一時的なコンピュータ可読媒体 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101552640B (zh) * | 2008-04-01 | 2012-04-11 | 富士通株式会社 | 滤波器系数变更装置和方法 |
US8515293B2 (en) * | 2009-05-07 | 2013-08-20 | Nec Corporation | Coherent receiver |
WO2012105070A1 (ja) * | 2011-02-01 | 2012-08-09 | 日本電気株式会社 | 光受信器、偏光分離装置および偏光分離方法 |
US9160459B2 (en) * | 2011-12-07 | 2015-10-13 | Nec Corporation | Equalization signal processor, optical receiver including the same, and method for equalization signal processing |
-
2020
- 2020-09-11 US US17/766,296 patent/US12009868B2/en active Active
- 2020-09-11 JP JP2021554153A patent/JP7327502B2/ja active Active
- 2020-09-11 WO PCT/JP2020/034526 patent/WO2021079647A1/ja active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012118215A1 (ja) * | 2011-03-02 | 2012-09-07 | 日本電気株式会社 | 光受信器、偏波分離装置、および光受信方法 |
WO2018198891A1 (ja) * | 2017-04-25 | 2018-11-01 | 日本電気株式会社 | 信号合成装置及び信号合成方法 |
WO2019198465A1 (ja) * | 2018-04-12 | 2019-10-17 | 日本電気株式会社 | 光空間通信システム、光受信装置、光受信方法、及び非一時的なコンピュータ可読媒体 |
Non-Patent Citations (2)
Title |
---|
CHO, J. ET AL.: "Probabilistic Constellation Shaping for Optical Fiber Communications", JOURNAL OF LIGHTWAVE TECHNOLOGY, vol. 37, no. 6, March 2019 (2019-03-01), pages 1590 - 1607, XP011717096, DOI: 10.1109/JLT.2019.2898855 * |
SAVORY, S. J. ET AL.: "Digital filters for coherent optical receivers", OPTICS EXPRESS, vol. 16, no. 2, January 2008 (2008-01-01), pages 804 - 817, XP007906766, DOI: 10.1364/OE.16.000804 * |
Also Published As
Publication number | Publication date |
---|---|
US12009868B2 (en) | 2024-06-11 |
US20240056195A1 (en) | 2024-02-15 |
JP7327502B2 (ja) | 2023-08-16 |
JPWO2021079647A1 (ja) | 2021-04-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CA3099841C (en) | Dimensional transformation in optical communication | |
US10587358B1 (en) | Probabilistic constellation shaping across time and frequency | |
Faruk et al. | Digital signal processing for coherent transceivers employing multilevel formats | |
JP6315040B2 (ja) | 光伝送歪補償装置、光伝送歪補償方法及び通信装置 | |
US8705977B1 (en) | Equalizer tap correction and reseeding | |
US10763972B2 (en) | Timing recovery for optical coherent receivers in the presence of polarization mode dispersion | |
US9300401B2 (en) | Updating apparatus and method for equalizer coefficient, receiver and otpical communication system | |
US20180034552A1 (en) | Signal processing device used in optical receiver and signal processing method | |
US9025651B1 (en) | Simplified polarization mode dispersion equalization | |
WO2017059783A1 (en) | Method and apparatus for residual phase noise compensation | |
US10938483B1 (en) | Training-aided feedforward channel equalization | |
US10164715B2 (en) | Adaptive demapper | |
WO2021079647A1 (ja) | 係数更新量出力装置、係数更新装置、係数更新量出力方法及び記録媒体 | |
Zhao et al. | Carrier phase recovery based on KL divergence in probabilistically shaped coherent systems | |
US11855699B2 (en) | Optical transmission system, optical transmitting apparatus and optical receiving apparatus | |
Savory | Digital signal processing for multilevel modulation formats | |
WO2013174304A1 (en) | Method with improved phase robustness in coherent detected optical system | |
Savory et al. | DSP for optical transponders | |
WO2023067641A1 (ja) | デジタル信号処理回路、方法、受信機、及び通信システム | |
US20240137127A1 (en) | Optical transmitter that transmits multi-level signal | |
WO2023284617A1 (en) | System and method for interference cancellation in optical transmission | |
Shu et al. | Experimental investigation of extended Kalman Filter combined with carrier phase recovery for 16-QAM system | |
Nguyen et al. | Carrier Phase Recovery for Optical Coherent M-QAM Communication Systems Using Harmonic Decompositionbased Maximum Loglikelihood Estimators |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20878497 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 17766296 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 2021554153 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20878497 Country of ref document: EP Kind code of ref document: A1 |