JP6767310B2 - Optical transmission / reception system - Google Patents

Description

The present invention relates to an optical transmission / reception system.

In the optical subscriber system, a PON (Passive Optical Network) system standardized by IEEE (The Institute of Electrical and Electronics Engineers, Inc.) and ITU-T (International Telecommunication Union Telecommunication Standardization Sector) is widely used. There is. In a PON system, each of a plurality of subscriber line termination devices (hereinafter referred to as "ONU" (Optical Network Unit)) is coupled to a single optical fiber via an optical splitter installed outside the accommodation station. It has a network configuration connected to a subscriber line terminal device (hereinafter referred to as "OLT" (Optical Line Terminal)) of an accommodation station.

In a PON system, an uplink signal and a downlink signal are transmitted in both directions on the same optical fiber with different wavelengths. The downlink signal is a continuous signal in which signals addressed to ONUs installed in each of a plurality of subscribers' homes are multiplexed using a time division multiplexing (hereinafter referred to as "TDM" (Time Division Multiplexing)) technology. The ONU extracts and receives the signal of the time slot assigned to the own device from the continuous signal branched by the optical splitter. The uplink signal is a burst signal intermittently transmitted by the ONU, and is combined by an optical splitter to become a TDM signal, which is received by the OLT. In this way, in the PON system, the optical fiber from the OLT of the accommodation station to the optical splitter and the OLT arranged in the accommodation station can be shared by a plurality of subscribers, so that a high-speed optical access service exceeding giga is economical. Can be provided to.

In a PON system that has already been used as a commercial system, expansion of allowable transmission line loss is considered to be one of the major issues. If the allowable transmission line loss can be increased, the number of accommodated subscribers can be increased by utilizing an optical splitter with a large number of branches, and the accommodated area can be expanded by extending the transmission distance. Is possible.

To solve this problem, an optical amplifier is placed in the OLT, and this optical amplifier is used as a post-amplifier to increase the output intensity of the transmitter, while this optical amplifier is used as a preamplifier to improve the sensitivity of the receiver. Therefore, a method for increasing the allowable transmission line loss has been widely studied. However, the effect of improving the reception sensitivity as a preamplifier is small compared to the increase in the output intensity of the optical amplifier as a post amplifier. Therefore, in order to obtain an allowable transmission line loss equivalent to that of the downlink signal, it is necessary to further improve the reception sensitivity of the uplink signal.

In order to improve the reception sensitivity of uplink signals, it is considered to apply digital coherent transmission technology to a PON system (see, for example, Non-Patent Document 1). By applying the receiving technique in the digital coherent transmission technique, the receiving sensitivity can be significantly improved as compared with the intensity modulation direct detection method which has been used in the PON system so far.

Further, the reception sensitivity can be further improved by 3 dB by using the phase modulated signal instead of the intensity modulated signal. In recent years, a quadrature phase-shifted signal, that is, a QPSK (Quadrature Phase Shift Keying) signal has been used in a large-capacity relay transmission system. The signal generation of QPSK is performed using a lithium niobate (LiNbO ₃ ) Machzenda modulator (hereinafter referred to as “LN modulator”). The LN modulator is a device suitable for a long-distance transmission system because it has excellent characteristics in terms of modulation band and coupling with an optical fiber. However, since this device has a length of several centimeters, it is not practical to apply it to a subscriber network, especially to an ONU.

As a compact modulator to replace the LN modulator, an indium phosphide (InP) semiconductor Machzenda modulator (hereinafter referred to as "InP modulator") is being developed and is beginning to be commercialized. In the InP modulator, the change in the refractive index of the material that causes the phase modulation is larger than that of LiNbO ₃ , so that the device can be manufactured in the length of millimeter order. Further, since the InP modulator is a semiconductor device, it can be monolithically integrated with optical components such as a semiconductor laser and a semiconductor optical amplifier (SOA), and the transmitter / receiver can be further miniaturized. However, the InP modulator currently has problems such as yield, and further cost reduction technology needs to be developed in order to apply it to an access system.

As a method of generating a phase modulation signal without using these expensive Machzenda modulators, there is a direct modulation method in which the drive current of a semiconductor laser is driven by a transmission signal. However, it is the frequency, not the phase, of the light output from the semiconductor laser that can be directly modulated by the transmitted signal. This frequency modulation is realized by setting the bias current of the semiconductor laser to a high value and driving the semiconductor laser in a region where the adiabatic chirp is dominant.

As an example, FIG. 14A shows an optical spectrum when a semiconductor laser is directly modulated by a binary digital baseband signal. As shown in FIG. 14A, the “1” signal (hereinafter referred to as “mark”) having a large drive current shifts to the high frequency side by Δf with respect to the center frequency (f ₀ ), and the “1” signal having a small drive current “ The "0" signal (hereinafter referred to as "space") shifts to the low frequency side by Δf. In reality, the output intensity of the semiconductor laser changes at the same time as the drive current of the semiconductor laser is changed. Therefore, as shown in FIG. 14B, the light intensity on the space side is slightly lower than that on the mark side. To do.

Data is transmitted by adjusting the amount of frequency change so that the phase change after a certain period of time caused by the frequency change becomes π or zero so that the signal can be received by the differential detection described later. At that time, when the phase difference from the previous bit is π or zero, it is necessary to differentially code the transmission signal in advance so as to receive a mark or a space, respectively. As described above, the signal generated by the direct modulation method in which the drive current of the semiconductor laser is driven by the transmission signal is a frequency modulation signal on the surface, but is a signal that can be used as a phase modulation signal.

As a method for generating a phase modulation signal using this frequency modulation, a method called continuous phase frequency shift keying (hereinafter referred to as "CPFSK" (Continuous Phase Frequency Shift Keying)) and a customary phase modulation, that is, PSK (Phase Shift Keying) There is a method called). In order to avoid confusion with the general phase modulation method that directly generates the phase modulation signal corresponding to the transmission signal by using the Mach Zenda modulator described above, the latter PSK is referred to as phase modulation by direct modulation in the following. That is, it is called DM-PSK (Direct Modulation-PSK).

The relationship between DM-PSK and CPFSK is shown in FIG. In FIG. 15, the horizontal axis represents the modulation index of frequency modulation (denoted as FM modulation index in FIG. 15), and the vertical axis represents the duty ratio. Hereinafter, in order to briefly show the difference between DM-PSK and CPFSK, a case of transmitting binary digital data will be described as an example. Assuming that the bit rate of the signal is B, the modulation index m is defined by the following equation (1).

16 and 17 show the operating principle of DM-PSK. As shown in FIG. 16, an optical pulse according to the duty ratio (D) is output in the mark section, and the modulation index of frequency modulation is such that the optical phase changes by π with respect to the previous bit at the end of the optical pulse. Alternatively, the intensity of the optical pulse is adjusted (see, for example, Non-Patent Document 2). Since the optical phase does not change after the end of the optical pulse, the optical phase change after 1 bit time elapses remains π. At this time, there is a relationship of the following equation (2) between the modulation index (m) and the duty ratio (D).

In the scheme shown in FIG. 16, the optical phase continues to increase over time. On the other hand, as shown in FIG. 17, the intensity of the optical pulse in the mark section may be output alternately with a bit increasing and a bit decreasing with respect to the intensity of the space section (for example, non-patent documents). 3). In the method shown in FIG. 17, the optical phase does not increase with the passage of time and becomes either zero or π. Here, assuming that the bit rate is B, the signal reception can be performed by differential detection using a signal sequence delayed by 1 bit time, that is, 1 / B in any of the methods shown in FIGS. 16 and 17. it can.

On the other hand, as shown in FIG. 15, in CPFSK, the duty ratio (D) is always 1 regardless of the modulation index (m). That is, in CPFSK, the optical phase change after 1 bit time elapses exceeds π when the modulation index (m) exceeds 0.5, and the delay time (T) required for differential detection is 1 bit time. That is, it is represented by the following equation (3) instead of 1 / B (see, for example, Non-Patent Document 4).

When the modulation index (m) is 0.5, a phase modulation signal that can be demodulated with the minimum frequency shift amount, that is, B / 4, is generated, and by 1 bit time, that is, 1 / B differential detection. Signal reception is performed. This case is particularly referred to as minimum shift keying (hereinafter referred to as "MSK" (Minimum Shift Keying)), and as shown in FIG. 15, it can also be referred to as DM-PSK having a duty ratio of 1.

As described above, the phase-modulated signal generated by direct modulation such as CPFSK and DM-PSK can receive the signal by differential detection. However, when coherent reception is performed by homodyne detection, the frequency difference between the signal light and the locally oscillated light (hereinafter referred to as "local emission") becomes a problem. Hereinafter, the case of CPFSK will be described in detail with reference to the mathematical formulas as an example. In the process shown by the following formula, each of the signals of the I (In-phase) component and the Q (Quadrature) component generated by homodyne detection using a 90-degree optical hybrid detector is AD (Analog to). Digital) This is a process performed on a signal obtained by sampling with a converter and digitizing it.

The sampling frequency is set to a value that is at least twice the bit rate B of the signal based on Shannon's sampling theorem. In the case of a multi-valued signal, B is the symbol rate. Generally, when a station emission with a center frequency of f ₀ and a signal light of f ₁ are incident on a 90-degree optical hybrid detector and homodyne detected, the I component and Q of the signal current output from the 90-degree optical hybrid detector are obtained. The components are represented by the following equations (4) and (5), respectively.

Equation (4), in (5), _{f m} is the frequency difference between the local light and the signal light _{(f m = f 0 -f 1} ), φ is the phase difference. Further, t is time, and A is a coefficient for associating the cosine function (cos) and the sine function (sin) with the photoelectric flow rate. When equations (4) and (5) are expressed in Euler by complex numbers, the following equation (6) is obtained.

In CPFSK, since the frequency shifts of Δf and −Δf are given to each of the mark and the space, the received signal current of the mark is represented by the equation (7), and the received signal current of the space is expressed by the equation (8). Shown.

Using equations (7) and (8), when transitioning from mark to mark, the output after differential detection is represented by the following equation (9).

Similarly, the output after differential detection when transitioning from mark to space is expressed by Eq. (10), and the output after differential detection when transitioning from space to space is expressed by Eq. (11). The output after differential detection in the case of transition from space to mark is expressed by Eq. (12).

Equations (7) and (8) match with each other having a phase of φ at t = 0. Therefore, the equations (10) and (12) show the case where the sign transitions while maintaining the phase continuity at t = 0, and the signal identification is performed at t = T. Substituting t = T into equations (10) and (12) gives equations (13) and (14). The value obtained by modifying the equation (1) was applied as Δf, and the value of the equation (3) was applied as T.

First, consider the case where _fm = 0. Equations (9) and (14) are cases where the mark is received, and the received signal converges at the position (0, A ² ) on the IQ plane regardless of the sign of the previous bit. Further, Eqs. (11) and (13) are cases where a space is received, and the received signal converges at the position (0, −A ² ) on the IQ plane regardless of the sign of the previous bit. ..

However, when the sign is inverted with respect to the previous bit, as is clear from equations (10) and (12), it moves on the circumference of the received signal constellation at angular velocities of 4πΔf and -4πΔf, respectively. To do. FIG. 18A is a diagram showing this situation.

Next, when f _m ≠ 0, as shown in FIG. 18 (b), the received signal constellation is phase-rotated by 2π f _m t on the IQ plane with respect to the case of FIG. 18 (a). ..

As a matter of fact, the received signal constellation after the differential detection is deteriorated due to the insufficient band of the semiconductor laser and the 90-degree optical hybrid detector used, so that the signal point is a predetermined signal point shown in FIG. 18 (a). The received signal does not converge at the positions 0, A ² ) and (0, -A ² ). In order to compensate for the deterioration of the received signal constellation, digital signal processing technology is used in digital coherent transmission.

CMA (Constant Modulus Algorithm) is a commonly used signal processing algorithm. When CMA is used, FIR (Finite Impulse Response) so that the amplitude of the received signal constellation converges to a predetermined value, for example, (0, A ² ) and (0, −A ² ) in FIG. 18 (a). The tap coefficient of the filter is adaptively feedback-controlled. By filtering the received signal with an FIR filter in which the optimum value of the tap coefficient obtained by the feedback control is substituted, it is possible to compensate, that is, adapt or equalize the band shortage of the device used.

However, the CMA is effective when the circumferential radius of the received signal constellation is reduced due to the influence of the band shortage of the received signal. For example, this corresponds to the case where a QPSK generated by using a Machzenda modulator is used as a transmission signal.

On the other hand, in CPFSK and DM-PSK, the circumferential radius of the received signal constellation does not decrease even if the band used of the device is insufficient. Even when the different codes that generate the highest frequency components appear alternately, the received signal does not reach the predetermined signal point shown in FIG. 18A, but remains on the same circumference. In this case, even if CMA is used, it is not possible to compensate for the lack of bandwidth of the device used.

Therefore, in the case of CPFSK or DM-PSK, it is necessary to apply the LMS (Least Mean Square) algorithm. When the LMS algorithm is used, the optimum value of the tap coefficient of the FIR filter is obtained by the least squares method with a predetermined signal point on the IQ plane as the target value. As a target value, it is common to use a training signal which is a fixed pattern signal periodically embedded in a transmission signal sequence.

Dayou Qian, Eduardo Mateo, Ming-Fang Huang, “A 105km Reach Fully Passive 10G-PON using a Novel Digital OLT”, Tu.1.B.2, ECOC (European Conference on Optical Communication) Technical Digest, 2012, OSA. “4 Gbitls, 233-km OPTICAL FIBRE TRANSMISSION EXPERIMENT USING NEWLY PROPOSED DIRECT-MODULATION PSK” R.S. VODHANEL, “5 Gbit / s Direct Optical DPSK Modulation of a 1530-nm DFB Laser”, IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 1, No. 8, AUGUST 1989, pp.218-220. Katsushi Iwashita, Takao Matsumoto, “Modulation and Detection Characteristics of Optical Continuous Phase FSK Transmission System”, JOURNAL OF LIGHTWAVE TECHNOLOGY. VOL. LT-5, NO. 4, APRIL 1987, pp.452-460.

However, when the LMS algorithm is used for CPFSK or DM-PSK, the phase rotation of the received signal constellation caused by the frequency difference between the signal light and the station emission described above becomes a problem. Without this phase rotation, in the example of FIG. 18A, the LMS algorithm may be operated as a signal point targeting (0, A ² ) or (0, −A ² ). On the other hand, when the received signal constellation is phase-rotated as shown in FIG. 18B, the distance from the target signal point changes according to the amount of phase rotation. Therefore, FIR using the LMS algorithm. There is a problem that the calculation of the optimum value of the tap coefficient of the filter becomes unstable.

In view of the above circumstances, an object of the present invention is to provide a technique capable of accurately calculating the optimum value.

One aspect of the present invention is an optical transmission / reception system including an optical transmitter for transmitting an optical signal and an optical receiver for receiving the optical signal. The optical transmitter has a semiconductor laser and transmits data. On the other hand, the semiconductor laser uses a transmission signal generated by adding a fixed pattern of a plurality of predetermined signal points and having a uniform or almost uniform appearance frequency of each signal point as a drive signal. The optical receiver includes a locally oscillating light source that outputs a locally oscillating light and a locally oscillating light source that receives and transmits the locally oscillating light signal. Based on the light, the frequency-modulated optical signal is homodyne-detected, and a 90-degree optical hybrid detector that outputs an electric signal of an in-phase component and an orthogonal component, and each of the electric signals of the in-phase component and the orthogonal component are sampled. Obtained by differential detection, an AD converter that converts the digital sampling signal of the in-phase component and the sampling signal of the orthogonal component, a differential detection unit that differentially detects the sampling signal of the in-phase component and the orthogonal component, and differential detection. A phase average calculation unit that calculates a phase average value based on a fixed pattern of the signal points included in the signal, and a reverse rotation of the phases of the sampling signals of the in-phase component and the orthogonal component based on the phase average value. It is an optical transmission / reception system including a phase rotating unit for making the signal.

One aspect of the present invention is the optical transmission / reception system, wherein the frequency modulator changes the transmission intensity in the opposite direction with respect to the change characteristic of the intensity of the optical signal output by the semiconductor laser with respect to the drive signal. The electric field absorption type light modulator having a characteristic is provided, and the electric field absorption type light modulator performs intensity modulation on the frequency modulation optical signal based on the transmission signal.

One aspect of the present invention is the above-mentioned optical transmission / reception system, in which the frequency modulator has a transmission intensity characteristic in the same direction with respect to a change characteristic of the intensity of an optical signal output by the semiconductor laser with respect to the drive signal. The electric field absorption type light modulator is provided, and the electric field absorption type optical modulator performs intensity modulation on the frequency-modulated optical signal based on a signal in which the polarity of the transmission signal is inverted.

One aspect of the present invention is the above-mentioned optical transmission / reception system, wherein the optical receiver has an FIR filter unit that performs filtering based on a tap coefficient, the in-phase component in which the phase rotation unit is rotated in the reverse direction, and the orthogonal component. Is applied to the FIR filter unit to obtain a demodulated signal, the position of the sampled signal included in the demodulated signal on the IQ plane, and a plurality of predetermined signals serving as predetermined convergence target values. The distance is calculated based on the position of the point on the IQ plane, and based on the calculated distance, the tap of the FIR filter unit is performed so that the sampling signal converges to the nearest predetermined signal point. It further includes a tap coefficient calculation unit for calculating the optimum value of the coefficient.

According to the present invention, the optimum value calculation can be performed with high accuracy.

It is a figure which shows the structure of the optical transmission / reception system 100 in 1st Embodiment of this invention. It is a block diagram which shows the structure of the optical receiver in 1st Embodiment. It is a figure which shows the frame structure of the transmission signal of the optical receiver in 1st Embodiment. It is a block diagram which shows the structure of the optical receiver in 1st Embodiment. It is a block diagram which shows the other structural example of the optical receiver in 1st Embodiment. It is a block diagram which shows the other structural example of the optical receiver in 1st Embodiment. It is a figure (the 1) which shows the experimental result in 1st Embodiment. It is a figure (the 2) which shows the experimental result in 1st Embodiment. It is a figure (the 3) which shows the experimental result in 1st Embodiment. It is a block diagram which shows the structure of the optical receiver in the 2nd Embodiment of this invention. It is a graph which shows the characteristic of the semiconductor laser of the optical receiver of the 2nd Embodiment. It is a graph which shows the characteristic of the EA modulator of the optical receiver of the 2nd Embodiment. It is a block diagram which shows the other structural example of the optical receiver of 2nd Embodiment. It is a figure which shows the optical spectrum at the time of direct modulation using a semiconductor laser. It is a figure which shows the relationship between CPFSK and DM-PSK. It is a figure (the 1) for demonstrating the operation principle of DM-PSK. It is a figure (the 2) for demonstrating the operation principle of DM-PSK. It is a figure which shows the received signal constellation.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
(First Embodiment)
FIG. 1 is a diagram showing a configuration of an optical transmission / reception system 100 according to the first embodiment of the present invention. In the following description, a case where the optical transmission / reception system 100 is applied to a PON system will be described as an example. The PON system is an aspect of an optical transmission system. The optical transmission / reception system 100 includes ONU1 and OLT2. Although FIG. 1 shows a configuration in which the optical transmission / reception system 100 includes one ONU1, the optical transmission / reception system 100 may include a plurality of ONU1s.

The ONU1 is installed, for example, in the home of a subscriber who receives a communication service. The ONU 1 includes an optical transmitter 10. The optical transmitter 10 transmits an optical signal.
The OLT 2 is installed, for example, in a containment station. The OLT 2 includes an optical receiver 20. The optical receiver 20 receives the optical signal transmitted from the optical transmitter 10.

FIG. 2 is a block diagram showing a configuration of the optical transmitter 10 according to the first embodiment. The optical transmitter 10 includes a frequency modulation unit 102.
The frequency modulation unit 102 includes a gate switch 1021 and a semiconductor laser 1022. Based on the transmission signal and the gate signal, the gate switch 1021 outputs the transmission signal in the section corresponding to the time section in which the gate signal is supplied as a burst signal to the semiconductor laser 1022.

Here, the transmission signal is a signal having a frame configuration shown in FIG. 3, and is input from the outside. The transmission signal is a continuous signal and is composed of a portion of the burst signal to be transmitted and a portion of a fixed idle pattern superimposed between the burst signals and not transmitted. The burst signal portion has a configuration in which a preamble (hereinafter referred to as “PA” (Preamble)) portion is added to a payload portion including transmission data.

The PA part is a pattern in which a predetermined fixed pattern of symbols of L signal points used when calculating the amount of phase rotation is repeated M times (hereinafter, this pattern is also referred to as a fixed pattern of L × M symbols). Includes. Further, the portion other than the fixed pattern of the L × M symbol of PA includes, for example, a fixed pattern such as a training signal described later. The part of the fixed idle pattern is completely filled with a predetermined pattern that does not include information.

As a fixed pattern of the symbols of the L signal points, for example, a pseudo-random code sequence (PRBS (Pseudo Random Bit Sequence)) having a code length of 2 ⁿ -1 (n is a natural number) is applied. Further, M is a number of repetitions appropriately determined so that the fixed pattern has a sufficient length, and if the fixed pattern of the symbols of the L signal points has a sufficient length, M = 1, i.e., no repetition is required.
Further, the gate signal is input from the outside. The time section in which the gate signal is supplied corresponds to the section in which the burst signal exists in FIG. The gate signal is intermittently supplied multiple times. Therefore, the gate switch 1021 intermittently outputs a plurality of burst signals to the semiconductor laser 1022.

The semiconductor laser 1022 uses the burst signal output by the gate switch 1021 as a drive signal, drives it in a region dominated by the adiabatic chirp to perform direct modulation, and generates and outputs a frequency-modulated optical signal obtained by frequency-modulating the burst signal. .. Here, the frequency-modulated optical signal is, for example, an optical signal modulated by the above-mentioned CPFSK or DM-PSK method.

FIG. 4 is a block diagram showing a configuration of the optical receiver 20 according to the first embodiment. The optical receiver 20 includes a local oscillation light source 21, a 90-degree optical hybrid detector 22, AD converters 23-1 to 23-2, and a digital signal processing unit 24. The local oscillation light source 21 outputs local light emission that interferes with the received signal light.

The 90 degree optical hybrid detector 22 includes a splitter 221-1 to 221-2, a π / 2 delayer 222, a coupler 223-1 to 223-2 and a balanced receiver 224-1 to 224-2. The splitter 221-1 receives the frequency-modulated optical signal of the burst signal transmitted by the optical transmitter 10 as the received signal light, branches the received received signal light, and outputs the received signal light to the couplers 223-1 and 223-2. The splitter 221-2 receives the local emission output from the local oscillation light source 21, branches the received local emission, and outputs the light to the coupler 223-1 and the π / 2 delay device 222. The π / 2 delayer 222 delays the local emission output by the splitter 221-2 by π / 2 and outputs it to the coupler 223-2.

The coupler 223-1 generates interference light by merging and interfering with the received signal light output by the splitter 221-1 and the station emission output by the splitter 221-2, and produces two interference lights. It branches into interference light and outputs. The coupler 223-2 generates interference light by merging and interfering with the received signal light output by the splitter 221-1 and the station emission delayed by π / 2 minutes output by the π / 2 delayer 222. , The generated interference light is branched into two interference lights and output.

The balanced receiver 224-1 converts the two interference lights output by the coupler 223-1 into an electric signal, and detects and outputs the difference between the converted electric signals as an in-phase component, that is, an I component. The balanced receiver 224-2 converts the two interference lights output by the coupler 223-2 into an electric signal, and detects and outputs the difference between the converted electric signals as an orthogonal component, that is, a Q component. The AD converter 23-1 samples the analog electric signal of the I component and outputs it to the digital signal processing unit 24 as a digital sampling signal. The AD converter 23-2 samples the analog electric signal of the Q component and outputs it to the digital signal processing unit 24 as a digital sampling signal.

The digital signal processing unit 24 takes the digital sampling signal output from the AD converter 23-1 and the AD converter 23-2 as an input, and demodulates the received signal using a numerical calculation algorithm.
The processing performed by the optical receiver 20 is the same as that of a general homodyne receiver used in digital coherent transmission, except for the processing performed by the digital signal processing unit 24. The digital signal processing unit 24 includes a differential detection unit 241, a phase average calculation unit 242, a phase rotation unit 243, a training signal detection unit 244, a tap coefficient calculation unit 245, and an FIR filter unit 246.

The differential detection unit 241 performs differential detection based on the digital sampling signal of the I component output by the AD converter 23-1 and the digital sampling signal of the Q component output by the AD converter 23-2. The burst signal is demodulated, and the PA of the demodulated burst signal is output to the phase average calculation unit 242. The phase average calculation unit 242 refers to the sample points of the fixed pattern of the symbols of L signal points of R times (however, R ≦ M) among the fixed patterns of the L × M symbols included in the PA. To calculate the phase average value.

In the phase rotation unit 243, the digital sampling signal of the I component output by the AD converter 23-1 and the AD converter 23-2 are used for the phase rotation amount corresponding to the phase average value calculated by the phase average calculation unit 242. The output digital sampling signal of the Q component is rotated in the reverse direction on the IQ plane.

The training signal detection unit 244 detects the training signal included in the PA of the sampling signals of the I component and the Q component rotated in the reverse direction by the phase rotation unit 243. The tap coefficient calculation unit 245 obtains a demodulated signal of the training signal by giving the training signal detected by the training signal detection unit 244 to the FIR filter unit 246. Further, the tap coefficient calculation unit 245 makes the position of the training signal included in the obtained demodulation signal converge to the position of a predetermined target value on the IQ plane predetermined for the training signal. In addition, feedback control is adaptively performed to calculate the optimum tap coefficient of the FIR filter unit 246. Further, the tap coefficient calculation unit 245 uses the LMS algorithm, that is, the least squares method, in the process of converging the training signal to the position of a predetermined target value on the IQ plane and obtaining the optimum value of the tap coefficient. ..

The FIR filter unit 246 filters and demodulates the sampling signals of the I component and the Q component based on the tap coefficient of the optimum value calculated by the tap coefficient calculation unit 245, and outputs the demodulated I component signal and the Q component signal. ..

Here, a method of calculating the phase average value by the phase average calculation unit 242 will be described. As described above, when the CPFSK or DM-PSK signal is differentially detected, the received signal constellation is phase-rotated due to the frequency difference between the station emission and the received signal light. Due to this phase rotation, the calculation of the optimum value of the tap coefficient of the FIR filter using the LMS algorithm becomes unstable.

On the other hand, the phase average calculation unit 242 uses the following equation for the sample points of the L × R symbols among the fixed patterns of the L × M symbols included in the PA output by the differential detection unit 241. 15) Alternatively, the calculation of the following equation (16) is performed. It should be noted that which of the equation (15) and the equation (16) is applied is predetermined for each optical receiver 20.

In equations (15) and (16), i indicates the i-th sample point in the L × R symbol, and Σ _i indicates that the sum of the sample points of the total number M of the calculation targets is calculated. Further, 2 [pi] f m _T represents the phase rotation amount of a received signal constellation on I-Q plane caused by the frequency difference local light and the reception signal light. Here, _{f m,} as described above, the center frequency of the local light _{f 0,} if the center frequency of the reception signal light and the _{f 1,} is represented by f _{m =} f 0 -f _1. T is the delay time required for the differential detection represented by the equation (3). φ _i indicates the phase of the i-th sample point on the IQ plane when there is no phase rotation.

That is, equation (15) is an equation for calculating the average value of the angle portion of the sample points marked with Euler, and equation (16) is an argument for the sum of the exponential function portions of the sample points marked with Euler. That is, it can be said to be an equation for calculating the angle portion.

Comparing equations (15) and (16) with equations (9) to (12), φ _i corresponds to π / 2, π / 2 + 4πΔft, −π / 2, −π / 2-4πΔft. You can see that The received signal constellation of these sample points is as shown in FIG. 18A described above. FIG. 18A is symmetrical with respect to the I-axis, and the phase on the IQ plane takes positive and negative values with the I-axis as the boundary. Therefore, the equations (15) and (16) The average value of the part related to φ _{i in} ) is zero. As a result, the values of the equations (15) and (16) are approximated to 2πf _m T, and the phase rotation amount of 2π f _m T can be obtained by calculating the φ _ave of the equations (15) and (16). it can.

Phase rotation section 243, based on the phase rotation amount 2 [pi] f m _T phase average calculating unit 242 calculates, by reverse rotation of the received signal constellation in FIG. 18 (b), the received signal shown in FIG. 18 (a) Konsuta The ration is obtained. This can solve the problem that the optimum value calculation of the tap coefficient of the FIR filter unit 246 using the LMS algorithm by the tap coefficient calculation unit 245 becomes unstable.

If the variation in the appearance frequency of the signal points included in the fixed pattern of the L × M symbol becomes large, the received signal constellation in FIG. 18A is not subject to the I-axis, so that the average value of the portion related to φ _i is obtained. Does not become zero, and an error occurs in the calculation of the phase rotation amount. Therefore, the frequency of appearance of the signal points included in the fixed pattern needs to be uniform or almost uniform. In the case of substantially uniform, the average value of the portion related to φ _i does not become zero, but it suffices if there is negligible error uniformity compared to the phase average value of 2 π f _m T. Further, when the transmission signal is a binary baseband signal, the mark ratio may be 1/2 or a pattern of signal points close to 1/2. Further, the phase average calculation unit 242 calculates the phase average value by using the R times of the fixed patterns of the L × M symbols, but the R value is the appearance frequency of the signal points included in the R times. Is defined to be uniform or nearly uniform.

Further, for example, when the above-mentioned pseudo-random code sequence having a code length of 2 ⁿ -1 is applied as a fixed pattern of symbols of L signal points, the mark rate is halved if the value of n, which is a natural number, is small. You may not get close. In this case, a method such as adding one zero to set the code length to 2 ⁿ may be applied. Further, as a fixed pattern of the symbols of L signal points, a pattern in which 1 and 0 are repeated may be applied.

Further, as the training signal detected from the PA by the training signal detection unit 244, a fixed pattern of the L × M symbol may be applied as a training signal, or a fixed pattern other than the fixed pattern of the L × M symbol may be applied as a training signal. You may. When the fixed pattern of the L × M symbol is used as the training signal, the entire PA portion may be used as the fixed pattern of the L × M symbol. Further, when a fixed pattern other than the fixed pattern of the L × M symbol is applied as the training signal, the training signal is included in the remaining section of the PA as described above.

(Experimental result)
7 and 8 show the results of signal processing of digital data offline. For this digital data, a binary CPFSK signal with a signal bit rate of 10.0 Gbit / s and a modulation index of 0.6 is experimentally generated, and each of the I component and Q component obtained by homodyne detection is 50 GS (GigaSimple). It is digital data obtained by sampling at / s.

7 (a) and 7 (b) are the cases where the frequencies of the station emission and the signal light are substantially the same, and are diagrams showing the output obtained by differential detection and the output to which the LMS algorithm is applied, respectively. As shown in FIG. 7B, the LMS algorithm operates stably, and the binary signal points are clearly separated and demodulation is normally performed.

On the other hand, FIGS. 8A and 8B are diagrams showing the differentially detected output and the output to which the LMS algorithm is applied when the frequencies of the station emission and the signal light are deviated by about 4 GHz. As shown in FIG. 8B, the operation of the LMS algorithm becomes unstable, the binary signal points cannot be separated, and demodulation fails.

On the other hand, FIG. 9 shows the result when the technique of the optical transmission / reception system 100 of the present embodiment is applied to the same digital data as in FIG. 8A. As shown in FIG. 9A, the phase rotation of the reception constellation after differential detection is appropriately compensated. Further, as shown in FIG. 9B, even in the output to which the LMS algorithm is applied, the binary signal points are clearly separated and demodulation is normally performed.

According to the optical transmission / reception system 100 according to the first embodiment configured as described above, in the optical transmitter 10, the frequency modulation unit 102 has a semiconductor laser 1022, and a plurality of predetermined frequencies are set for the transmission data. The frequency-modulated light is driven by driving the semiconductor laser 1022 using a transmission signal generated by adding a fixed pattern in which the appearance frequency of each signal point is uniform or almost uniform. A signal is generated and a frequency modulated optical signal corresponding to the time interval indicated by the gate signal is transmitted. In the optical receiver 20, the local oscillation light source 21 outputs local light emission. The 90-degree optical hybrid detector 22 receives the frequency-modulated optical signal, homodyne detects the frequency-modulated optical signal based on the station emission, and outputs an electric signal having an in-phase component and an orthogonal component. Each of the AD converters 23-1 and 23-2 samples each of the electric signals of the in-phase component and the orthogonal component and converts them into a digital sampling signal of the in-phase component and a sampling signal of the orthogonal component. In the digital signal processing unit 24, the differential detection unit 241 differentially detects the sampling signals of the in-phase component and the orthogonal component. The phase average calculation unit 242 calculates the phase average value based on the fixed pattern of the signal points included in the signal obtained by the differential detection. The phase rotation unit 243 reversely rotates the phases of the sampling signals of the in-phase component and the orthogonal component based on the phase mean value.

With the above configuration, even when there is a frequency difference between the station emission and the frequency-modulated optical signal and phase rotation occurs, the appearance frequency of each signal point is uniform or almost uniform in a fixed pattern. By calculating the phase average value based on this, the phase component of each signal point on the IQ plane becomes zero, and the phase average value represents the amount of phase rotation. Therefore, by rotating the phase in the reverse direction based on the calculated phase rotation amount, the position of the received signal constellation can be returned to the position when there is no difference between the frequencies of the received signal light and the station emission. Thereby, an appropriate optimum value of the tap coefficient of the FIR filter unit 246 using the LMS algorithm by the tap coefficient calculation unit 245 can be calculated. Therefore, when the frequency-modulated optical signal generated by directly modulating the semiconductor laser 1022 is homodyne-detected using a 90-degree optical hybrid detector, the received signal constellation at the time of differential detection is the signal light and local emission. Due to the frequency difference of, it is possible to receive a signal even when the phase is rotated on the IQ plane.

<Modification example>
The optical transmitter 10 in the first embodiment may be configured as shown in FIG. FIG. 5 is a block diagram showing a configuration of an optical transmitter 10a applied in place of the optical transmitter 10 in the first embodiment. In the optical transmitter 10a, the same configurations as those of the optical transmitter 10 are designated by the same reference numerals, and different configurations will be described below.

The optical transmitter 10a includes a frequency modulation unit 102a. The frequency modulation unit 102a includes a semiconductor laser 1022 and a semiconductor optical amplifier 1023. The semiconductor optical amplifier 1023 receives the gate signal, changes the drive current by the gate signal to increase or decrease the gain, and corresponds to the time interval indicated by the gate signal in the continuous frequency-modulated optical signal output by the semiconductor laser 1022. Outputs a frequency-modulated optical signal. The time section in which the gate signal is supplied corresponds to the section in which the burst signal exists in FIG. Since the gate signal is intermittently supplied a plurality of times, the semiconductor optical amplifier intermittently outputs a plurality of burst signals.

In the optical transmitter 10 shown in FIG. 2, since the semiconductor laser 1022 is driven and directly modulated by using the burst signal as a drive signal, the output wavelength of the semiconductor laser 1022 may drift significantly at the rising edge of the burst signal. On the other hand, in the optical transmitter 10a shown in FIG. 5, since the semiconductor laser 1022 is not driven by the burst signal but is driven by the continuous transmission signal, the output wavelength of the semiconductor laser 1022 is stable. Therefore, the optical transmitter 10a is suitable for multiplexing optical signals at high-density wavelength intervals.

Further, the optical receiver 20 in the first embodiment may be configured as shown in FIG. FIG. 6 is a block diagram showing a configuration of an optical receiver 20a applied in place of the optical receiver 20 in the first embodiment. In the optical receiver 20a, the same configurations as those of the optical receiver 20 are designated by the same reference numerals, and different configurations will be described below. The optical receiver 20a includes a local oscillation light source 21, a 90-degree optical hybrid detector 22, AD converters 23-1 to 23-2, and a digital signal processing unit 24a. The digital signal processing unit 24a includes a differential detection unit 241, a phase average calculation unit 242, a phase rotation unit 243, a tap coefficient calculation unit 245a, and an FIR filter unit 246.
The tap coefficient calculation unit 245a calculates the optimum value of the tap coefficient by converging a predetermined signal point on the IQ plane as a target value.

The tap coefficient calculation unit 245a obtains a demodulated signal of the sampling signal by giving the sampling signals of the I component and the Q component rotated in the reverse direction by the phase rotation unit 243 to the FIR filter unit 246. The tap coefficient calculation unit 245a identifies the position of the sampling signal included in the demodulated signal on the IQ plane, that is, which signal point the sample point targets, so that the sample on the IQ plane is sampled. The distance between the position of the point and the position of a plurality of predetermined signal points is calculated.

Based on the calculated distance, the tap coefficient calculation unit 245a selects a predetermined signal point at the shortest distance to the position of the sample point as the position of the target value of the sample point. The tap coefficient calculation unit 245a applies the LMS algorithm, that is, the least squares method, and adaptively feedback-controls the sample points so that the sample points converge to the position of the selected target value, and the tap coefficient of the FIR filter unit 246 is controlled. Calculate the optimum value.

In the optical receiver 20 shown in FIG. 4, the tap coefficient calculation unit 245 calculates the optimum value of the tap coefficient of the FIR filter unit 246 based on the training signal detected by the training signal detection unit 244. Since it is known at which position on the IQ plane the training signal converges, the optimum value can be calculated in a relatively short time. On the other hand, since the time position accuracy required for training detection by the training signal detection unit 244 is very high, a large amount of resources are required for the signal processing circuit in the digital signal processing unit 24 to realize this accuracy. On the other hand, the tap coefficient calculation unit 245a of the optical receiver 20a calculates the optimum value of the tap coefficient without using the training signal. In this method, the time required to calculate the optimum value increases by the amount of distance calculation, but since the training signal is not used, the scale of the signal processing circuit in the digital signal processing unit 24a is adjusted to the scale of the digital signal processing unit of the optical receiver 20. It can be smaller than 24.

The combination of the optical transmitters 10 and 10a included in the ONU1 and the optical receivers 20 and 20a included in the OLT 2 is, for example, an optical transmitter 10 and an optical receiver 20, an optical transmitter 10 and an optical receiver 20a, and an optical transmitter 10a. The optical transmitter / receiver system 100 may be configured by any combination of the optical receiver 20, the optical transmitter 10a, and the optical receiver 20a.

(Second Embodiment)
In the second embodiment, the configuration of the optical transmitter among the configurations of the optical transmission / reception system is different from that of the optical transmission / reception system 100 in the first embodiment.
FIG. 10 is a block diagram showing the configuration of the optical transmitter 10b according to the second embodiment of the present invention. The optical transmitter 10b includes a frequency modulation unit 102b. The frequency modulation unit 102b includes a semiconductor laser 1022, a semiconductor optical amplifier 1023, and an electric field absorption type modulator 1024. The optical transmitter 10b is a configuration in which an electric field absorption type modulator 1024 is newly provided in addition to the configuration of the optical transmitter 10a of the first embodiment shown in FIG. The same configurations as those of the optical transmitter 10a are designated by the same reference numerals, and different configurations will be described below. In the following description, the electric field absorption type modulator 1024 will be referred to as an EA (Electro Absorption) modulator 1024.

In the optical transmitter 10 and the optical transmitter 10a, when the semiconductor laser 1022 directly modulates according to the drive current, not only the output optical carrier frequency but also the intensity is modulated at the same time. Therefore, the ideal frequency modulation as shown in FIG. 14A is not performed, and the intensity difference between the mark that shifts to the high frequency side as shown in FIG. 14B and the space that shifts to the low frequency side. Will occur.

In order to accurately determine the phase rotation amount of the received constellation according to the above equations (15) and (16), the sample points must exist on the circumference of the same radius on the IQ plane. is there. Therefore, if the difference in strength between the mark and the space shown in FIG. 14B is large, an error will occur in the calculation of the phase rotation amount.

In the optical transmitter 10b shown in FIG. 10, the difference in light intensity is made uniform by using the EA modulator 1024. In the optical transmitter 10b, the same transmission signal, that is, a binary digital baseband signal is input to the semiconductor laser 1022 and the EA modulator 1024 from the outside. The EA modulator 1024 performs intensity modulation on the frequency-modulated optical signal output by the semiconductor laser 1022 using the input transmission signal as a control signal.

FIG. 11 is a graph showing the relationship between the current value of the drive signal given to the semiconductor laser 1022 (hereinafter referred to as “drive current”) and the output light intensity. If an electric resistance is provided in the part where the drive current is applied, it is possible to drive with a voltage instead of the current. Therefore, the case where the horizontal axis is the drive voltage is also described.

As shown in FIG. 11A, when the drive current increases in the positive direction, the output light intensity of the semiconductor laser 1022 increases. Further, as shown in FIG. 11B, when the drive current increases in the negative direction, the output light intensity of the semiconductor laser 1022 increases. Therefore, in any case, it can be seen that if the binary digital baseband signal is directly modulated, an unnecessary intensity modulation component is generated together with the frequency modulation. In other words, when the drive current shown in FIG. 11A increases in the positive direction, an intensity modulation component having characteristics in the same direction as the characteristics of the transmitted signal is generated. When the drive current shown in FIG. 11B increases in the negative direction, the characteristics are inverted and an intensity modulation component having characteristics in the opposite direction to the characteristics of the transmitted signal is generated.

FIG. 12 is a graph showing the relationship between the voltage of the control signal and the transmission intensity characteristic in the EA modulator 1024. As shown in FIG. 12A, when the voltage of the control signal increases in the positive direction, the transmission intensity characteristic of the EA modulator 1024 decreases. On the other hand, as shown in FIG. 12B, when the voltage of the control signal increases in the negative direction, the transmission intensity characteristic of the EA modulator 1024 decreases. In other words, when the voltage of the control signal shown in FIG. 12A increases in the positive direction, intensity modulation in which the applied signal voltage and the sign are inverted is performed. When the voltage of the control signal shown in FIG. 12B increases in the negative direction, intensity modulation having the same sign as the applied signal voltage is performed.

Therefore, by combining the semiconductor laser 1022 having the characteristics of FIG. 11 (a) and the EA modulator 1024 having the characteristics of FIG. 12 (a), unnecessary intensity modulation components generated in the semiconductor laser 1022 are transferred to the EA modulator 1024. Can be removed. Further, the same effect can be obtained by combining the semiconductor laser 1022 having the characteristics shown in FIG. 11B and the EA modulator 1024 having the characteristics shown in FIG. 12B.

According to the optical transmitter / receiver in the second embodiment configured as described above, the optical transmitter 10b transmits in the opposite direction to the change characteristic of the intensity of the optical signal output by the semiconductor laser 1022 with respect to the drive signal. The EA modulator 1024 having an intensity change characteristic is provided, and the EA modulator 1024 performs intensity modulation on the frequency-modulated optical signal based on the input transmission signal.

There is also a method of making the intensity of marks and spaces uniform by using an optical filter such as a dielectric multilayer film, but since a member different from a semiconductor is used, this method requires miniaturization of optical components by monolithic integration. I can't. On the other hand, in the optical transmitters 10b and 10c of the second embodiment, the light intensity of the mark and the space is made uniform by using the EA modulator 1024, which is an optical component capable of monolithic integration, so that the light intensity is compact and economical. Depending on the configuration, the reception sensitivity of the modulated signal can be improved. Further, the configuration of the optical transmitters 10b and 10c of the second embodiment requires the same number of optical devices as compared with the case where the InP modulator is used, but the EA modulator 1024 is a general-purpose EA modulator. It is an optical device and is currently much cheaper than an InP modulator.

<Modification example>
The optical transmitter 10b in the second embodiment may be configured as shown in FIG. FIG. 13 is a block diagram showing a configuration of an optical transmitter 10c applied in place of the optical transmitter 10b in the second embodiment. The optical transmitter 10c includes a frequency modulation unit 102c. The frequency modulation unit 102c includes a semiconductor laser 1022, a semiconductor optical amplifier 1023, and an EA modulator 1024 (in FIG. 13, the electric field absorption type modulator 1024). The optical transmitter 10c is different from the configuration of the optical transmitter 10b shown in FIG. 10 in that the control signal input to the EA modulator 1024 is a control signal in which the polarity of the transmission signal is reversed. The same configurations as those of the optical transmitter 10b are designated by the same reference numerals, and different configurations will be described below. A transmission signal, that is, a binary digital baseband signal is input to the semiconductor laser 1022 from the outside, and a control signal in which the polarity of the transmission signal is reversed is input to the EA modulator 1024 from the outside.

The optical transmitter 10c includes an EA modulator 1024 having a characteristic of changing the transmission intensity in the same direction with respect to a characteristic of changing the intensity of the optical signal output by the semiconductor laser 1022 with respect to the drive signal, and the EA modulator 1024 is , Intensity modulation is performed on the frequency-modulated optical signal based on the signal in which the polarity of the input transmission signal is inverted. As a result, the intensity modulation component generated when the semiconductor laser 1022 is directly modulated can be removed by the EA modulator 1024.

The combination of the characteristics of the semiconductor laser 1022 and the EA modulator 1024 applied in the optical transmitter 10c is the semiconductor laser 1022 having the characteristics shown in FIG. 11 (a) and the EA modulator 1024 having the characteristics shown in FIG. 12 (b). Will be combined. Further, as another combination, the same effect can be obtained by combining the semiconductor laser 1022 having the characteristics shown in FIG. 11B and the EA modulator 1024 having the characteristics shown in FIG. 12A.

Further, in general, in a system that handles high-speed signals, digital baseband signals having different polarities, that is, differential signals are often used, and in such a case, it is preferable to apply the optical transmitter 10c.

Further, in the optical transmitters 10b and 10c, although the semiconductor optical amplifier 1023 is connected after the EA modulator 1024, it may be connected between the semiconductor laser 1022 and the EA modulator 1024.

<Modification example common to the first embodiment and the second embodiment>
In the first and second embodiments described above, an example in which a binary digital baseband signal is mainly applied as a transmission signal is described, but the configuration of the present invention may be limited to the embodiment. Absent. A digital baseband signal having an N value (N is an integer of 3 or more) may be used to generate an N value CPFSK signal or a DM-PSK signal.
In the first and second embodiments described above, the configuration in which the optical transmission / reception system 100 is applied to the PON system is shown, but the optical transmission / reception system 100 may be applied to other optical transmission systems.

10, 10a, 10b, 10c ... Optical transmitter, 102, 102a, 102b, 102c ... Frequency modulator, 1021 ... Gate switch, 1022 ... Semiconductor laser, 1023 ... Semiconductor optical amplifier, 1024 ... Electric field absorption type modulator, 20, 20a ... Optical receiver, 21 ... Local oscillation light source, 22 ... 90 degree optical hybrid detector, 221-1, 221-2 ... Splitter, 222 ... π / 2 delayer, 223-1, 223-2 ... Coupler, 224 -1, 224-2 ... Balanced receiver, 23-1, 23-2 ... AD converter, 24 ... Digital signal processing unit, 241 ... Differential detector, 242 ... Phase average calculation unit, 243 ... Phase rotation unit , 244 ... Training signal detection unit, 245, 245a ... Tap coefficient calculation unit, 246 ... FIR filter unit

Claims (4)

An optical transmission / reception system including an optical transmitter for transmitting an optical signal and an optical receiver for receiving the optical signal.
The optical transmitter
Transmission generated by having a semiconductor laser and adding a fixed pattern of a plurality of predetermined signal points to the transmission data in which the appearance frequency of each signal point is uniform or almost uniform. A frequency modulation unit that drives the semiconductor laser using a signal as a drive signal to generate and transmit a frequency-modulated optical signal.
With
The optical receiver
A locally oscillating light source that outputs locally oscillating light,
A 90-degree optical hybrid detector that receives the frequency-modulated optical signal, homodyne-detects the frequency-modulated optical signal based on the locally oscillated light, and outputs an electric signal having an in-phase component and an orthogonal component.
An AD converter that samples each of the electrical signals of the in-phase component and the orthogonal component and converts them into a digital sampling signal of the in-phase component and a sampling signal of the orthogonal component.
A differential detection unit that differentially detects sampling signals of the in-phase component and the orthogonal component,
A phase average calculation unit that calculates a phase average value based on a fixed pattern of the signal points included in the signal obtained by differential detection, and a phase average calculation unit.
A phase rotating unit that reversely rotates the phases of the sampling signals of the in-phase component and the orthogonal component based on the phase mean value.
Optical transmission / reception system. The frequency modulator
An electric field absorption type optical modulator having a change characteristic of transmission intensity in the opposite direction with respect to a change characteristic of the intensity of an optical signal output by the semiconductor laser with respect to the drive signal is provided.
The optical transmission / reception system according to claim 1, wherein the electric field absorption type optical modulator performs intensity modulation on the frequency-modulated optical signal based on the transmission signal. The frequency modulator includes an electric field absorption type optical modulator having a transmission intensity characteristic in the same direction with respect to a change characteristic of the intensity of an optical signal output by the semiconductor laser with respect to the drive signal.
The optical transmission / reception system according to claim 1, wherein the electric field absorption type optical modulator performs intensity modulation on the frequency-modulated optical signal based on a signal in which the polarity of the transmission signal is inverted. The optical receiver
FIR filter unit that performs filtering based on tap coefficient,
A sampling signal of the in-phase component and the orthogonal component in which the phase rotating section is rotated in the reverse direction is given to the FIR filter section to obtain a demodulating signal, and the position of the sampling signal included in the demodulating signal on the IQ plane , The distance is calculated based on the positions of a plurality of predetermined signal points serving as predetermined convergence target values on the IQ plane, and the sampling signal is the closest based on the calculated distance. A tap coefficient calculation unit that calculates the optimum value of the tap coefficient of the FIR filter unit so as to converge to a predetermined signal point, and a tap coefficient calculation unit.
The optical transmission / reception system according to any one of claims 1 to 3, further comprising.

Images