JP5645632B2 - Optical transceiver and optical transmission / reception method - Google Patents

Description

  The present invention relates to an optical transceiver and an optical transmission / reception method used in digital communication using an optical fiber.

  A 40 Gb / s DBPSK (Differential Binary Phase-Shift Keying) signal has a wide occupied band and is vulnerable to spectrum constriction in ROADM (Reconfigurable Optical Add-drop Multiplexer). Therefore, it has been difficult to realize a frequency interval of 50 GHz at the time of wavelength multiplexing. In addition, the symbol rate is high and the fiber dispersion characteristics such as chromatic dispersion and polarization mode dispersion are weak.

  As a technique for solving these problems, there is a so-called Partial DPSK (Partial Differential Phase-Shift Keying) system (for example, see Patent Document 1). The Partial DPSK scheme is a scheme in which a free spectral range (FSR) of a one-symbol delay interferometer provided in a DPSK receiver is sufficiently wide with respect to a symbol rate 1 / T (T is a symbol period).

  And this Partial DPSK system has a full width at half maximum (FWHM) (Full Width Half Maximum) of less than 1 / T in the optical region or a full width at half maximum (FWHM) of 0. This is a method capable of obtaining good reception characteristics even when band narrowing is performed so as to be less than 5 / T.

  This Partial DPSK system is particularly effective for the DBPSK system that takes signal points only on the I-axis (In-phase axis) in the complex electric field plane. In the complex electric field plane, the DQPSK system, which takes signal points on both the I axis and the Q axis (Quadrature-phase axis), is not as resistant to the band limitation in the optical region as the DBPSK method, but is limited in the band in the electric region. On the other hand, it has the same strength as the DBPSK method.

  A DA converter (DAC: Digital-to-Analog Converter) applicable to digital signal processing (DSP: Digital Signal Processing) of optical fiber communication has been developed with a sampling rate exceeding 20 Gsample / s ( For example, refer nonpatent literature 1).

  The increase in sampling rate is very severely limited by tradeoffs with circuit scale, power consumption, waveform distortion, and the like. When waveform generation is performed using a DA converter, an oversampling ratio (= sampling rate to symbol rate ratio) of 2 times is basically required. This is to prevent waveform distortion due to aliasing that is widely known in the field of wireless communication.

  FIG. 4 is a block diagram showing an example (1) of the optical transmission system according to the prior art. Specifically, in a conventional optical transmitter, when performing digital arithmetic processing and DA conversion with double oversampling, a block diagram focusing on discrete time-continuous time conversion, and a signal of each block processed The change of the frequency spectrum is shown.

  First, it is assumed that there is an analog signal (continuous time expression) S1 that is to be expressed by pre-equalization processing. For digital arithmetic processing (DSP), the frequency spectrum of the complex baseband signal S3a obtained by synthesizing the I-ch signal and the Q-ch signal when converted to discrete time by the double oversampling unit 12a is 2 / T intervals. Will be repeated. For this reason, the I-ch and Q-ch real-time electrical signals are expressed as 2 as a signal S2a in which the main component of the signal S1 is extracted in advance by a low-pass filter (LPF) 11a having a 1 / T band. It is necessary to give to the double oversampling unit 12a.

  When the signal S3a is converted to the continuous-time signal S4a by the same double oversampling DA converter 13a, the frequency spectrum is also repeated at 2 / T intervals. For this reason, after the DA converter 13a, it is necessary to limit the band by the low-pass filter 14a having a band of about 1 / T to obtain the signal S5a. Thus, it can be seen that a signal having a substantially desired spectrum can be realized up to the first null points of the signals S1, S2a, S3a, S4a, and S5a.

  FIG. 5 is a block diagram showing an example (2) of the optical transmission system according to the prior art. Specifically, in a conventional optical transmitter, when performing digital arithmetic processing and DA conversion with 1-time oversampling, a block diagram focusing on discrete time-continuous time conversion, and the signal subjected to processing of each block The change of the frequency spectrum is shown. However, the bands of the low-pass filters 11a and 14a are the same as those in FIG.

  Also in the case of the configuration of FIG. 5, if there is an analog signal (continuous time expression) S1 to be expressed by pre-equalization processing, the signal S1 passes through the same low-pass filter 11a as in FIG. Is converted into discrete time. The frequency spectrum of the complex baseband signal obtained by synthesizing the I-ch signal and the Q-ch signal after the discrete time is repeated at 1 / T intervals. For this reason, it is understood that the spectrum is overlapped in the digital domain as indicated by the signal S3b.

  It can be seen that when the I-ch and Q-ch real-time electrical signals are extracted by a low-pass filter having a 1 / T band, the spectrum overlaps in the digital domain. When converting to a continuous time signal by the DA converter 13b subsequent to the 1 × oversampling unit 12b, the frequency spectrum is also repeated at 1 / T intervals. For this reason, it can be seen that aliasing can no longer be removed and a desired spectrum cannot be realized by using any low-pass filter including the low-pass filter 14a having a 1 / T limit band in the subsequent stage of the DA converter 13b. .

  FIG. 6 is a block diagram showing an example (3) of the optical transmission system according to the prior art. Specifically, in a conventional optical transmitter, when performing digital arithmetic processing and DA conversion with 1-time oversampling, a block diagram focusing on discrete time-continuous time conversion, and the signal subjected to processing of each block The change of the frequency spectrum is shown. However, in the case of the configuration of FIG. 6, the limited bands of the low-pass filters 11b and 14b are set to half (0.5 / T) of the case of FIG.

  In the case of the configuration of FIG. 6, if there is an analog signal (continuous time expression) S1 that is desired to be expressed by pre-equalization processing, the signal S1 passes through the low-pass filter 11b and is then converted to discrete time by the 1 × oversampling unit 12b. Is done. The frequency spectrum of the complex baseband signal S3c obtained by synthesizing the I-ch signal and the Q-ch signal after the discrete time is repeated at 1 / T intervals as shown in FIG. Therefore, the main components of the I-ch and Q-ch real-time electrical signals need to be extracted in advance by the low-pass filter 11b having a bandwidth of 0.5 / T.

  When the subsequent DA converter 13b of the 1 × oversampling unit 12b converts the signal into a continuous time signal, the frequency spectrum is repeated at 1 / T intervals. For this reason, it is necessary to limit the band by the low-pass filter 14b having a band of about 0.5 / T in the subsequent stage of the DA converter 12b. As a result, it can be seen that a signal whose band is strictly narrowed to about half the symbol rate is generated. In this case, aliasing does not occur.

  In recent years, the bit rate per wavelength in optical fiber communication reaches 40 Gb / s and 100 Gb / s. Therefore, it is difficult to perform double oversampling using a 20 Gsample / s DA converter due to the operating speed limit of the DA converter.

  On the other hand, a method has been studied in which the oversampling ratio is set to 1 and waveform generation is performed by a DA converter, and antialiasing is prevented by an antialiasing filter at the transmission end and the reception end (see Non-Patent Document 2, for example).

US Patent Application Publication No. 2007/0196110 (Japanese Patent Publication No. 2009-534993)

P. Schvan et al. "A 22GS / s 6b DAC with integrated digital ramp generator" ISSCC 2005, 6.7 (2005). L. Dou et al. , “Electronic pre-distortion operating at (1) sample / symbol with accumulating bias control for CD compensation” OFC / NFOEC 2010, OThT4 (2010)

However, the above prior art has the following problems.
In a conventional optical transceiver, wavelength dispersion and frequency spectrum narrowing are pre-equalized by digital signal processing at the transmission end, an analog optical waveform is generated, and asynchronous detection is performed at the reception end. However, it has been difficult to realize the above method with a low power consumption and a small circuit scale by using an available DA converter at a bit rate of 40 Gb / s or more.

  Further, in the method disclosed in Non-Patent Document 2, there is a problem that waveform distortion remains and reception characteristics deteriorate unless coherent detection is performed at the receiving end and equalization by digital signal processing is combined at the receiving end.

  In order to avoid such reception characteristic deterioration, assuming double oversampling, a DA converter requires a very high sampling rate at a bit rate of 40 Gb / s or more. As a result, there are problems in terms of hardware such as power consumption, circuit scale, waveform quality, feasibility, cost, and availability.

  The present invention has been made in order to solve the above-described problems. Even at a bit rate of 40 Gb / s or more, low power consumption and a small-scale circuit can be realized without increasing the oversampling ratio. An object is to obtain an optical transceiver and an optical transmission / reception method excellent in hardware.

The optical transceiver according to the present invention generates an analog signal by performing digital signal processing and DA conversion processing at a 1-time oversampling ratio on a digital signal having a sampling rate of 20 Gsample / s or more. And an optical transmitter that modulates the light source output and sends the modulated optical signal through the optical transmission line, and the free spectral interval for the modulated optical signal received through the optical transmission line. 2 / T (where T is one symbol time) to give a delay amount, generate an addition component and a subtraction component of delayed light and non-delayed light, and based on the addition component and subtraction component, 0.5 An optical receiver that generates an electric signal band-limited at / T, and the optical transmitter performs pre-equalization processing for suppressing intersymbol interference of the transmission signal and has a band limit of approximately 0.5 / T A digital signal processing unit that prevents aliasing by a low-pass filter, has a sampling rate of 20 Gsample / s or more, and discretizes the oversampling ratio corresponding to the sampling rate to symbol rate ratio as 1 Sample / Symbol, and the sampling rate and oversampling In order to prevent aliasing of a digital-analog converter having a sampling ratio and converting a digital signal discretized by a digital signal processing unit into an analog signal and an analog signal converted by the digital-analog converter An analog low-pass filter that limits the low-frequency band of the optical modulator, and a modulator that modulates the light source output based on the output of the analog low-pass filter and transmits the modulated optical signal via the optical transmission path, Through the optical transmission line A delay interferometer that gives a delay amount having a free spectral interval of approximately 2 / T to the modulated optical signal received in step S1 and generates an addition component and a subtraction component of delayed light and non-delayed light; A balanced photon detector that has an electrical band of 5 / T and generates an electrical signal based on an addition component and a subtraction component generated by a delay interferometer, and an electrical signal generated by the balanced photon detector A data identification unit that performs parallel development and data identification of electrical signals based on the regenerated clock and outputs an electrical signal after parallel development and data identification is included.

An optical transmission / reception method according to the present invention is an optical transmission / reception method using an optical transmitter and an optical receiver, and in the optical transmitter , digital signal processing and digital signal processing for a digital signal having a sampling rate of 20 Gsample / s or more. An optical transmission step of generating an analog signal by performing DA conversion processing at an oversampling ratio of 1 ×, modulating a light source output based on the analog signal, and transmitting the modulated optical signal via an optical transmission path; In the optical receiver, the modulated optical signal received via the optical transmission line is given a delay amount by setting the free spectral interval to approximately 2 / T (where T is one symbol time), and the delayed light An optical reception step that generates an addition component and a subtraction component of the non-delayed light and generates an electrical signal band-limited at 0.5 / T based on the addition component and the subtraction component. The optical transmission step performs pre-equalization processing for suppressing intersymbol interference of the transmission signal and prevents aliasing by a low-pass filter having a limit band of approximately 0.5 / T, and is 20 Gsample / s or more. A digital signal processing step for discretizing the oversampling ratio corresponding to the sampling rate to symbol rate ratio as 1 Sample / Symbol, and having a sampling rate and an oversampling ratio and being discrete at the digital signal processing step A digital-to-analog conversion step for converting the digitized digital signal into an analog signal, a filtering step for performing low-frequency band limitation for preventing aliasing on the analog signal after conversion by the digital-to-analog conversion step, A modulation step of modulating the light source output based on the output of the tapping step and transmitting the modulated optical signal via the optical transmission line, and the optical reception step includes the modulated light received via the optical transmission line A delay interference step for giving a delay amount having a free spectral interval of approximately 2 / T to a signal, generating an addition component and a subtraction component of delayed light and non-delayed light, and an electrical band of approximately 0.5 / T An electrical signal generation step for generating an electrical signal based on the addition component and the subtraction component generated by the delay interference step, and a clock reproduced based on the electrical signal generated in the electrical signal generation step And a data identification step of performing parallel development and data identification of the electric signal and outputting an electric signal after the parallel development and data identification .

  According to the optical transceiver and the optical transmission / reception method of the present invention, digital signal processing and DA conversion processing are performed at an oversampling ratio of 1 at the transmission end, and the free spectral interval of the delay interferometer is approximately 2 at the reception end. Even if the bit rate is 40 Gb / s or higher by reducing the bandwidth narrowing penalty in the electrical domain at the transmitting end by setting the bandwidth to 0.5 / T and setting the balanced photon detector to 0.5 / T. Therefore, it is possible to obtain an optical transceiver and an optical transmission / reception method excellent in hardware that can realize low power consumption and a small-scale circuit without increasing the oversampling ratio.

It is a block diagram which shows the optical transmitter-receiver which concerns on Embodiment 1 of this invention. It is explanatory drawing which shows the simulation result (1) of the optical transmitter-receiver which concerns on Embodiment 1 of this invention. It is explanatory drawing which shows the simulation result (2) of the optical transmitter-receiver which concerns on Embodiment 1 of this invention. It is a block diagram which shows the example (1) of the form of the optical transmission system which concerns on a prior art. It is a block diagram which shows the example (2) of the form of the optical transmission system which concerns on a prior art. It is a block diagram which shows the example (3) of the form of the optical transmission system which concerns on a prior art.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of an optical transceiver and an optical transmission / reception method according to the invention will be described with reference to the drawings.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing an optical transceiver according to Embodiment 1 of the present invention, where FIG. 1 (a) shows an optical transmitter and FIG. 1 (b) shows an optical receiver. .

  The optical transmitter includes a multiplexing unit (MUX) 101, an encoding unit (ENC) 102, a digital signal processing unit (DSP) 103, DA converters (DACs) 104-1 and 104-2, and drivers 105-1 and 105-. 2. Low-pass filters (LPF) 106-1 and 106-2, a light source (TLD) 107, and a modulator 108 are configured to output a modulated optical signal via the optical transmission line 120-1. The

  On the other hand, the optical receiver includes a delay interferometer (DLI: Delay Interferometer) 130, balanced photon detectors (Balanced Receivers) 131-1 and 131-2, and a clock regeneration / demultiplexing unit 132. The output signal from the optical transmitter is received via the optical transmission line 120-2, signal processing is performed, and an electric signal after parallel development and data identification described later is output.

Next, the connection relationship of each component will be described.
In the optical transmitter, a multiplexing unit 101, an encoding unit 102, and a digital signal processing unit 103 are connected in series to an input signal. Further, the digital signal processing unit 103 is connected to DA converters 104-1 and 104-2 on I-ch and Q-ch, and further passes through drivers 105-1 and 105-2, respectively. And 106-2.

  The low pass filters 106-1 and 106-2 are connected to the modulator 108. The modulator 108 is configured to modulate the optical signal from the light source 107 and output the modulated optical signal to the optical transmission line 120-1.

  On the other hand, in the optical receiver, a delay interferometer 130 is connected to the optical transmission line 120-2. The delay interferometer 130 is connected to balanced photon detectors 131-1 and 131-2 on I-ch and Q-ch, respectively. Further, the outputs of the balanced photon detectors 131-1 and 131-2 are sent to the clock recovery / demultiplexing unit 132 for data identification.

  Next, the operation of the optical transmitter shown in FIG. Here, an input signal with a bit rate of 40 Gb / s is assumed, but the present invention is not limited to this. The modulation scheme of the modulator 108 uses the DQPSK scheme, but may be a DP (Dual-polarized) -DBPSK scheme or a DP-DQPSK scheme.

  First, the multiplexing unit 101 multiplexes 16 parallel SFI-5 (Serdes Framer Interface Level 5) signals input from outside (not shown) at 2.5 Gb / s to generate a 40 Gb / s signal, and an encoding unit 102 Output to.

  The encoding unit 102 differentially encodes the 40 Gb / s signal input from the multiplexing unit 101 for the DQPSK modulation method in the modulator 108, and outputs the 40 Gb / s signal after differential encoding to the digital signal processing unit 103. Output.

  For example, the digital signal processing unit 103 pre-equalizes chromatic dispersion and nonlinearity of the transmission line optical fiber, and performs anti-aliasing filtering using the low-pass filters 106-1 and 106-2 having a bandwidth equivalent to 0.5 / T. Then, the I-ch component of the filtered digital signal is output to the DA converter 104-1, and the Q-ch component is output to the DA converter 104-2.

  The DA converter 104-1 converts the I-ch digital signal input from the digital signal processing unit 103 into an analog signal and outputs the analog signal to the driver 105-1. On the other hand, the DA converter 104-2 converts the Q-ch digital signal input from the digital signal processing unit 103 into an analog signal and outputs the analog signal to the driver 105-2.

  The driver 105-1 amplifies the I-ch analog signal input from the DA converter 104-1 to an amplitude necessary for driving the modulator 108, and outputs the amplified signal to the low-pass filter 106-1. On the other hand, the driver 105-2 amplifies the Q-ch analog signal input from the DA converter 104-2 to an amplitude necessary for driving the modulator 108, and outputs the amplified signal to the low-pass filter 106-2.

  The low-pass filter 106-1 performs a low-frequency band limitation of approximately full width at half maximum (FWHM) = 1 / T to prevent aliasing of the I-ch amplified signal input from the driver 105-1, and the I-ch band limitation The signal is output to the modulator 108 as a signal. On the other hand, the low-pass filter 106-2 performs a low-frequency band limitation on the full width at half maximum (FWHM) = 1 / T in order to prevent aliasing from being applied to the Q-ch amplified signal input from the driver 105-2. The band-limited signal is output to the modulator 108.

  The light source 107 generates unmodulated light and outputs it to the modulator 108. As the light source, for example, light with a C band of 1550.2 nm and +16 dBm may be generated using variable wavelength.

  The modulator 108 modulates the optical signal input from the light source 107 using the I-ch band limited signal input from the low pass filter 106-1 and the Q-ch band limited signal input from the low pass filter 106-2. To the optical transmission line 120-1. As the modulator, for example, a two-parallel Mach-Zehnder modulator may be used.

  The optical transmission line 120-1 includes an optical amplifier, a multiplexing / demultiplexing device, a transmission line optical fiber, a relay device, and the like, and transmits the optical signal output from the modulator 108 to the optical receiver.

Next, the operation of the optical receiver shown in FIG.
The optical transmission line 120-2 includes an optical amplifier, a multiplexing / demultiplexing device, a transmission line optical fiber, a relay device, and the like. For example, the optical transmission line 120-2 transmits the optical signal output from the above-described optical transmitter and supplies the optical signal to the delay interferometer 130. .

  The delay interferometer 130 branches the optical signal input from the optical transmission path 120-2 into an I-ch pre-delay optical signal and a Q-ch pre-delay optical signal.

  The optical signal before delayed interference for I-ch is divided into light (delayed light) given a delay amount T1 and light (non-delayed light) given a delay amount T2. Then, the delay interferometer 130 generates “π / 4 phase-shifted delayed light” in which the optical phase of the delayed light is shifted by π / 4, and an addition component of this π / 4 phase-shifted delayed light and non-delayed light (I-ch delayed interfering signal addition component) and subtracting component (I-ch delayed interfering signal subtraction component) are generated. However, T1> T2. Here, the inverse 1 / T3 of T3 = T1−T2 corresponds to the free spectral interval (FSR) of the delay interferometer 130.

  Similarly, the optical signal before delay interference for Q-ch is divided into light (delayed light) given a delay amount T1 and light (non-delayed light) given a delay amount T2. The delay interferometer 130 generates “−π / 4 phase-shifted delayed light” in which the optical phase of the delayed light is shifted by −π / 4, and the −π / 4 phase-shifted delayed light and the non-delayed light. An addition component (signal addition component after Q-ch delay interference) and a subtraction component (signal subtraction component after Q-ch delay interference) are generated. Also for Q-ch, the free spectral interval (FSR) is 1 / T3.

  In addition, when generating the addition component and subtraction component of delayed light and non-delayed light, the phase difference of one ch (corresponding to I-ch here) of I-ch and Q-ch is π. It is important that the phase difference of the other channel (corresponding to Q-ch here) is shifted by -π / 4, which is opposite in sign to the phase difference of one channel, as described above. It is not limited to shifting the phase of light by π / 4.

  The delay interferometer 130 adds an addition component (an I-ch post-delay delayed signal addition component) and a subtraction component (I-ch post-delay interfering signal subtraction component) generated from the I-ch pre-delay optical signal. Addition component (Q-ch post-delay interference post-interference signal addition component) and subtraction component (Q-ch post-delay interference post-interference signal) output from the balanced photon detector 131-1 and generated from the optical signal before delay interference for Q-ch Subtraction component) is applied to the balanced photon detector 131-2.

  The balanced photon detector 131-1 square-detects each of the I-ch delayed interfering signal addition component and the I-ch post-interfering signal subtraction component input from the delay interferometer 130, converts the result into an electric signal, and then calculates the difference. Take. Furthermore, the balanced photon detector 131-1 generates an electric signal (signal after I-ch detection) that has been subjected to current-voltage conversion and amplification for the difference result, and outputs it to the clock recovery / demultiplexing unit 132.

  The balanced photon detector 131-2 square-detects the signal addition component after Q-ch delay interference and the signal subtraction component after I-ch delay interference input from the delay interferometer 130, converts the result into an electric signal, and then performs a difference. Take. Furthermore, the balanced photon detector 131-2 generates an electric signal (signal after Q-ch detection) that has been subjected to current-voltage conversion and amplification with respect to the difference result, and outputs it to the clock recovery / demultiplexing unit 132.

  The clock regeneration / demultiplexing unit 132 is based on the post-I-ch detection signal output from the balanced photon detector 131-1 and the post-Q-ch detection signal output from the balanced photon detector 131-2. Play the clock component. Further, the clock recovery / demultiplexing unit 132 performs data identification of the signal after I-ch detection and the signal after Q-ch detection based on the recovered clock.

  In the clock recovery / demultiplexing unit 132, data identification is performed in parallel development (or parallel development after data identification), and an electric signal after parallel development and data identification is output to the outside (not shown).

  In the configuration of FIG. 1 according to the first embodiment, the low-pass filters 106-1 and 106-2 are arranged immediately after the drivers 105-1 and 105-2, but the drivers 105-1 and 105- are shown. You may arrange | position just before 2. Further, if the band limitation of the drivers 105-1 and 105-2 serves as an anti-aliasing filter (that is, if unnecessary components can be removed), the low-pass filters 106-1 and 106-2 may be removed.

  Next, simulation results performed to obtain optimum specifications of each element in the optical transceiver according to the first embodiment having the configuration shown in FIG. 1 will be described with reference to FIGS. FIG. 2 is an explanatory diagram showing a simulation result (1) of the optical transceiver according to the first embodiment of the present invention. FIG. 3 is an explanatory diagram showing a simulation result (2) of the optical transceiver according to the first embodiment of the present invention.

First, the calculation conditions of FIG. 2 are shown below.
-Bit rate of input signal of optical transmitter: 40 Gb / s
Modulation method of modulator 108: DQPSK
-Tap length of FIR (Finite Impulse Response) filter for chromatic dispersion pre-equalization included in the digital signal processing unit 103: 17 symbols-Oversampling of the digital signal processing unit 103 and the DA converters 104-1 and 104-2 Ratio: 1 time Resolution of DA converters 104-1 and 104-2: 6 bits Full width at half maximum (FWHM) of optical filter: 20 GHz
-Electric bandwidth limit after DA conversion: 12 GHz
-Electric band limitation of optical receiver: 12 GHz (equivalent to band limitation of balanced photon detectors 131-1 and 131-2 at approximately 0.5 / T)
Free spectral interval (FSR) of delay interferometer 130 of optical receiver: 1 / T

  In the simulation of FIG. 2, regarding the chromatic dispersion of the transmission line 120-2, two types of 0 ps / nm or 400 ps / nm are used. As for pre-equalization, digital signal processing is used to completely equalize chromatic dispersion pre-equalization (described as “pre-equalization” in FIG. 2), or not equalized at all (in FIG. 2). , Described as “No pre-equalization”). In addition, pre-equalization for the optical band limitation is not performed.

  In the case of zero chromatic dispersion (0 ps / nm) shown in FIGS. 2 (a1) and (a2), it can be seen that the eye is closed more than half irrespective of the presence or absence of pre-equalization. In addition, in the case of the chromatic dispersion of 400 ps / nm shown in FIGS. 2 (b1) and (b2), it can be seen that the eye is completely closed regardless of the presence or absence of pre-equalization.

  On the other hand, the calculation conditions of FIG. 3 are the same as the calculation conditions of FIG. 2 except that the free spectral interval (FSR) of the delay interferometer 130 of the optical receiver is set to 1.8 / T instead of 1 / T. It is. Further, the simulation was performed with respect to the chromatic dispersion and pre-equalization of the transmission line 120-2 in the same manner as in FIG. Further, pre-equalization for the optical band limitation is not performed as in FIG.

  In the case of zero chromatic dispersion (0 ps / nm) shown in FIGS. 3A1 and 3A2, it can be seen that a good eye opening is obtained. In the case of chromatic dispersion 400 ps / nm, there is a penalty if there is no pre-equalization as shown in FIG. 2 (b2), but if chromatic dispersion pre-equalization is performed as shown in FIG. 2 (b1). You can see that the eyes open.

  This simulation result is not an optimization of the tap coefficient of the FIR filter. Therefore, a good eye is obtained by optimizing the tap coefficient using an algorithm such as MMSE (Minimum Mean Square Error) known in wireless communication. An opening will be obtained.

  Therefore, as can be seen from a comparison between FIG. 2 showing the simulation result at FSR = 1 / T and FIG. 3 showing the simulation result at FSR = 1.8 / T, the free spectral interval of the delay interferometer is By setting it to approximately 2 / T, it is possible to reduce the band narrowing penalty in the electrical region of the transmitting end.

  As described above, according to the first embodiment, digital signal processing and DA conversion processing are performed at a 1 × oversampling ratio at the transmission end. Furthermore, at the receiving end, instead of severely limiting the band in the electrical domain by the anti-aliasing filter, the free spectral interval of the delay interferometer is set to approximately 2 / T, and the balanced photon detector is approximately 0.5 / T. Bandwidth is limited.

  As a result, the bandwidth narrowing penalty in the electrical domain at the transmitting end can be reduced, and even at a bit rate of 40 Gb / s or higher, low power consumption and a small-scale circuit can be realized without increasing the oversampling ratio. An optical transceiver and an optical transmission / reception method excellent in hardware can be obtained.

  Furthermore, the band limitation in the electrical domain of the DQPSK method can be solved by setting the free spectral interval sufficiently larger than the symbol rate. For this reason, as a modulation method, in addition to the DBPSK method and the DP-DBPSK method, a DQPSK method can also be adopted.

  Furthermore, the symbol rate can be reduced by improving the multilevel by the DP-DBPSK method or the DQPSK method. For this reason, it contributes also to the reduction of the sampling rate calculated | required by DA converter.

  101 Multiplexer (MUX), 102 Encoder (ENC), 103 Digital Signal Processor (DSP), 104-1, 104-2 DA Converter (DAC), 105-1, 105-2 Driver, 106-1 , 106-2 low-pass filter (LPF), 107 light source (TLD), 108 optical modulator, 120-1, 120-2 optical transmission line, 130 delay interferometer (DLI), 131-1, 131-2 balanced photon Detector, 132 clock recovery / demultiplexing unit (DEMUX).

Claims (4)

A digital signal having a sampling rate of 20 Gsample / s or more is subjected to digital signal processing and DA conversion processing at an oversampling ratio of 1 time to generate an analog signal, and a light source output is modulated based on the analog signal. An optical transmitter for sending the modulated optical signal through an optical transmission line;
For the modulated optical signal received via the optical transmission path, a free spectral interval is set to approximately 2 / T (where T is one symbol time) to give a delay amount, and delayed light and non-delayed An optical receiver that generates an addition component and a subtraction component with light, and generates an electric signal band-limited at approximately 0.5 / T based on the addition component and the subtraction component ;
The optical transmitter is
A pre-equalization process for suppressing intersymbol interference of a transmission signal is performed, aliasing is prevented by a low-pass filter whose band is approximately 0.5 / T, a sampling rate of 20 Gsample / s or more is provided, and a sampling rate pair A digital signal processing unit that discretizes the oversampling ratio corresponding to the symbol rate ratio as 1 Sample / Symbol;
A digital-analog converter having the sampling rate and the oversampling ratio, and converting the digital signal discretized by the digital signal processing unit into an analog signal;
An analog low-pass filter that performs low-frequency band limitation to prevent aliasing with respect to the analog signal after conversion by the digital-analog converter;
A modulator that modulates the light source output based on the output of the analog low-pass filter, and sends the modulated optical signal via an optical transmission line;
Including
The optical receiver is:
A delay that gives a delay amount having a free spectral interval of approximately 2 / T to the modulated optical signal received via the optical transmission line, and generates an addition component and a subtraction component of delayed light and non-delayed light. An interferometer,
A balanced photon detector having an electrical band of approximately 0.5 / T and generating an electrical signal based on the addition component and the subtraction component generated by the delay interferometer;
Data that performs parallel development and data identification of the electrical signal based on the clock regenerated based on the electrical signal generated by the balanced photon detector, and outputs the electrical signal after parallel development and data identification With the identification part
An optical transceiver characterized by comprising: The optical transceiver according to claim 1 ,
The optical transmitter / receiver characterized in that the modulator in the optical transmitter has a modulation scheme of polarization multiplexed differential four phase shift keying (DP-DQPSK). The optical transceiver according to claim 1 ,
The optical transmitter / receiver is characterized in that the modulator in the optical transmitter has a modulation scheme of polarization multiplexed differential two phase shift keying (DP-DBPSK). An optical transmission / reception method using an optical transmitter and an optical receiver,
In the optical transmitter, an analog signal is generated by performing digital signal processing and DA conversion processing at a 1 × oversampling ratio on a digital signal having a sampling rate of 20 Gsample / s or more, and based on the analog signal An optical transmission step of modulating the light source output and transmitting the modulated optical signal via an optical transmission path;
In the optical receiver, for the modulated optical signal received via the optical transmission line, a free spectral interval is set to approximately 2 / T (where T is one symbol time) to give a delay amount. An optical reception step of generating an addition component and a subtraction component of delayed light and non-delayed light, and generating an electric signal band-limited at approximately 0.5 / T based on the addition component and the subtraction component; Prepared ,
The optical transmission step includes
A pre-equalization process for suppressing intersymbol interference of a transmission signal is performed, aliasing is prevented by a low-pass filter whose band is approximately 0.5 / T, a sampling rate of 20 Gsample / s or more is provided, and a sampling rate pair A digital signal processing step of discretizing the oversampling ratio corresponding to the symbol rate ratio as 1 Sample / Symbol;
A digital-to-analog conversion step of converting the digital signal having the sampling rate and the oversampling ratio and discretized in the digital signal processing step into an analog signal;
A filtering step for performing low-frequency band limitation for preventing aliasing with respect to the analog signal after the conversion by the digital-analog conversion step;
A modulation step of modulating the light source output based on the output of the filtering step, and transmitting the modulated optical signal via an optical transmission line;
Including
The optical receiving step includes
A delay that gives a delay amount having a free spectral interval of approximately 2 / T to the modulated optical signal received via the optical transmission line, and generates an addition component and a subtraction component of delayed light and non-delayed light. An interference step;
An electrical signal generating step having an electrical band of approximately 0.5 / T and generating an electrical signal based on the addition component and the subtraction component generated by the delay interference step;
Data identification for performing parallel development and data identification of the electric signal based on the clock regenerated based on the electric signal generated in the electric signal generation step, and outputting the electric signal after parallel development and data identification Step and
Light transmission and reception method, which comprises a.

Images