Optical transmitter

Abstract

An optical transmitter (6000) according to one embodiment includes: a modulation unit (6400) that includes two data modulation units that modulate the data of each of optical pulse strings bifurcated from an optical CSRZ pulse string in which the frequency of an electric-field waveform is B / 2, an optical coupling unit that couples output optical signals from the two data modulation units, and an optical delay unit that gives a delay of time τ to the optical pulse string that goes through one of the two data modulation units; and a digital data processing unit (6100) that includes a sub-signal generation unit (6120) that separates the symbol string of the optical signal to be ultimately transmitted from an optical modulation unit, alternately for each symbol, and outputs two trains of sub-signals u1 and u2, a first sample shift unit (6131) that performs a sample shift equivalent to a 1 / (2B) delay on u2 and outputs u2', a 2 x 2 computation unit (6140) that obtains v1 and v2 from u1 and u2' by equation (1), and a second sample shift unit (6132) that performs a sample shift corresponding to B on v2' and obtains v2.

Description

[Technical Field]

[0001] This disclosure relates to an optical transmitter for optical fiber communications. [Background technology]

[0002] Time-division multiplexing (TDM) is one method for achieving high-speed communication in optical fiber communication systems. In particular, the CSRZ-TDM method, which uses CSRZ (Carrier-suppressed-return-to-zero) pulses, is known as a method that can output a signal with twice the symbol rate of the drive signal while suppressing redundant spectral spread (Patent Document 1, Non-Patent Document 1).

[0003] Figure 1(a) shows an optical modulator for generating a CSRZ-TDM signal in the first reference example of Patent Document 1. The optical modulator 1000 shown in Figure 1(a) comprises an input optical port 1001, a pulse generation unit 1111, an optical branching unit 1231, a first data modulation unit 1221, a second data modulation unit 1222, an optical delay unit 1241, an optical multiplexing unit 1232, and an output optical port 1002. The optical paths in which the first data modulation unit 1221 and the second data modulation unit 1222 are arranged are designated as arm 1 and arm 2, respectively.

[0004] The pulse generation unit 1111 is a push-pull driven Mach-Zehnder modulation circuit that is biased to a null point and driven by a clock of frequency B / 2 output from the clock source 1151.

[0005] The optical delay unit 1241 delays the CSRZ pulse train passing through arm 2 by τ relative to the CSRZ pulse train passing through arm 1. In this disclosure, when simply referred to as "delay," it refers to group delay and not phase delay. Here, the optical delay unit 1241 is designed such that τ = 1 / (2B).

[0006] The waveform (envelope excluding the carrier component) of the output optical field E of the pulse generation unit 1111 can be represented as a sine wave with frequency B / 2, if harmonic components are ignored for simplicity. The peak interval, i.e., the pulse interval, as seen in the intensity waveform is 1 / B, and the sign of the optical phase is reversed between adjacent pulses. The pulse train input to the second data modulation unit 1222 lags the pulse train input to the first data modulation unit 1221 by τ = 1 / (2B) (i.e., 1 / 4 period). Therefore, if the symbol timing in the first data modulation unit 1221 is set to the position marked with a triangle (▽) in Figure 1(b), and the symbol timing in the second data modulation unit 1222 is set to the position marked with a triangle (△) in Figure 1(c), then even after the outputs of each data modulation unit 1221 and 1222 are combined in the optical multiplexer 1232, no interference occurs in the symbol timing, and a CSRZ-TDM signal with a symbol rate of 2B is obtained. Note that Figures 1(b) and (c) show the electric field waveform excluding the optical carrier component and the data modulation component.

[0007] Various optical modulation circuits can be used as data modulation units 1221 and 1222, such as optical intensity modulation circuits, optical IQ modulation circuits, and polarization multiplexed optical IQ modulation circuits. Although the optical modulator 1000 may be integrally formed on a single optical waveguide substrate, if it is considered separately as an optical CSRZ pulse generation unit 1100 consisting of a pulse generation unit 1111 and a clock source 1151, and an optical modulation unit 1200 consisting of the remaining parts, equivalent functionality can be obtained by forming the optical circuit portion of the optical CSRZ pulse generation unit 1100 and the optical modulation unit 1200 on separate optical waveguide substrates and connecting them with an optical fiber.

[0008] For CSRZ-TDM, an optical modulator such as the optical modulator 2000 shown in Figure 2(a) is also possible. The input optical port 2001, pulse generation unit 2111, optical branching unit 2231, first data modulation unit 2221, second data modulation unit 2222, optical delay unit 2241, optical multiplexing unit 2232, and output optical port 2002 that constitute the optical modulator 2000 shown in Figure 2(a) correspond to the input optical port 1001, pulse generation unit 1111, optical branching unit 1231, first data modulation unit 1221, second data modulation unit 1222, optical delay unit 1241, optical multiplexing unit 1232, and output optical port 1002 that constitute the optical modulator 1000 shown in Figure 1(a).

[0009] The only difference between the optical modulator 2000 shown in Figure 2(a) and the optical modulator 1000 shown in Figure 1(a) is that the optical delay unit 2241 is located after the data modulation unit 2222, rather than before. In this case, if the symbol timing during modulation in the data modulation units 2221 and 2222 is the same as the symbol timing shown in Figures 1(b) and (c), it is obvious that a signal equivalent to that of the optical modulator 1000 shown in Figure 1 can be obtained. That is, since the optical delay unit 2241 imparts a delay of τ = 1 / (2B) to the CSRZ pulse train passing through the arm 2 after data modulation, there is no interference in symbol timing even after the combined signals are combined in the optical multiplexer 2232, and a CSRZ-TDM signal can be obtained.

[0010] Compared to the optical modulator 1000 shown in Figure 1, the optical modulator 2000 shown in Figure 2 has an implementation advantage in that the high-frequency electrical wiring from the drive signal output source to the optical modulator 2000 can be of equal length. Similar to the optical modulator 1000, the optical modulator 2000 can achieve the same functionality whether the optical circuit portion of the optical CSRZ pulse generation unit 2100 and the optical modulation unit 2200 are integrally formed on a single optical waveguide substrate, or whether they are formed on separate optical waveguide substrates and connected by optical fibers.

[0011] To configure an optical transmitter that outputs a CSRZ-TDM signal using optical modulators as shown in Figures 1 and 2, the signal source and drive system should be configured so that the symbol sequence of the signal to be ultimately output is alternately distributed between the first and second data modulation sections.

[0012] In the optical modulator 1000 shown in Figure 1, when polarization multiplexed IQ modulation units are used as data modulation units 1221 and 1222, the layout can be as shown in Figure 3. The optical modulation unit 3200 shown in Figure 3 comprises an optical branching unit 3231, an optical delay unit 3241, a first polarization multiplexed IQ modulation unit 3221, a second polarization multiplexed IQ modulation unit 3222, an X-polarization side optical multiplexing unit 3232-X, and a Y-polarization side optical multiplexing unit 3232-Y.

[0013] The first polarization multiplexed IQ modulation unit 3221 comprises an optical branching unit 3261, IQ modulation units 3221-X and 3221-Y, a polarization 90-degree rotation unit 3270, and a polarization coupling unit 3280. The second polarization multiplexed IQ modulation unit 3222 comprises an optical branching unit 3262, IQ modulation units 3222-X and 3222-Y, a polarization 90-degree rotation unit 3270, and a polarization coupling unit 3280. That is, the polarization 90-degree rotation unit 3270 and the polarization coupling unit 3280 are shared between the first polarization multiplexed IQ modulation unit 3221 and the second polarization multiplexed IQ modulation unit 3222. This is because the polarization 90-degree rotation unit 3270 and the polarization coupling unit 3280 are generally larger than other circuit elements and are often technically difficult to manufacture. Instead, the function corresponding to the optical multiplexer 1232 in Figure 1 is decomposed into the X-polarization side optical multiplexer 3232-X and the Y-polarization side optical multiplexer 3232-Y. For the IQ modulation sections 3221-X, 3221-Y, 3222-X, and 3222-Y, a typical nested Mach-Zehnder configuration IQ modulation section is used. Furthermore, in Figure 3, the positions of the IQ modulation sections 3221-Y and 3222-X may be swapped, and the waveguide intersection may be placed on the input side of the IQ modulation section instead of the output side.

[0014] Furthermore, in the optical modulator 2000 shown in Figure 2, when polarization multiplexed IQ modulation units are used as data modulation units 2221 and 2222, the layout can be as shown in Figure 4. The optical modulation unit 4200 shown in Figure 4 comprises an optical branching unit 4231, an X-polarization side optical delay unit 4241-X, a Y-polarization side optical delay unit 4241-Y, a first polarization multiplexed IQ modulation unit 4221, a second polarization multiplexed IQ modulation unit 4222, an X-polarization side optical multiplexing unit 4232-X, and a Y-polarization side optical multiplexing unit 4232-Y.

[0015] The first polarization multiplexed IQ modulation unit 4221 comprises an optical branching unit 4261, an X-polarization side IQ modulation unit 4221-X, a Y-polarization side IQ modulation unit 4221-Y, a polarization 90-degree rotation unit 4270, and a polarization coupling unit 4280. The second polarization multiplexed IQ modulation unit 4222 comprises an optical branching unit 4262, an X-polarization side IQ modulation unit 4222-X, a Y-polarization side IQ modulation unit 4222-Y, a polarization 90-degree rotation unit 4270, and a polarization coupling unit 4280. That is, the polarization 90-degree rotation unit 4270 and the polarization coupling unit 4280 are shared by the first polarization multiplexed IQ modulation unit 4221 and the second polarization multiplexed IQ modulation unit 4222. The reason is the same as in the case of the optical modulation unit 3200 shown in Figure 3. Instead, the function corresponding to the optical multiplexer 2232 in Figure 2 is decomposed into the X-polarization side optical multiplexer 4232-X and the Y-polarization side optical multiplexer 4232-Y, and further, the function corresponding to the optical delayer 2241 in Figure 2 is decomposed into the X-polarization side optical delayer 4241-X and the Y-polarization side optical delayer 4241-Y. Note that the X-polarization side IQ modulationers 4221-X, 4222-X and the Y-polarization side IQ modulationers 4221-Y, 4222-Y are each IQ modulationers with a typical nested Mach-Zehnder configuration. Also, in Figure 4, the positions of the Y-polarization side IQ modulationer 4221-Y and the X-polarization side IQ modulationer 4222-X may be swapped, and the waveguide crossover may be placed on the input side instead of the output side of the IQ modulationer.

[0016] Figure 5 shows the configuration of a conventional optical transmitter 5000, which has a trivial configuration as an optical transmitter that outputs a CSRZ-TDM signal. The optical transmitter 5000 comprises a digital signal processing unit 5100, a transmitting laser 5200, an optical CSRZ pulse generation unit 5300, an optical modulation unit 5400, delay adjustment units 5511 and 5512, and drivers 5521 and 5522.

[0017] The optical CSRZ pulse generator 5300 corresponds to the clock source 1151 and pulse generator 1111 of the optical modulator 1000 shown in Figure 1, or the clock source 2151 and pulse generator 2111 of the optical modulator 2000 shown in Figure 2. The optical modulation unit 5400 corresponds to the portion from the optical branching unit 1231 to the optical multiplexing unit 1232 of the optical modulator 1000 shown in Figure 1, or the portion from the optical branching unit 2231 to the optical multiplexing unit 2232 of the optical modulator 2000 shown in Figure 2. As the data modulation units 1221 and 1222 in Figure 1, or the data modulation units 2221 and 2222 in Figure 2, polarization-multiplexed optical IQ modulation units laid out as the optical modulation unit 3200 shown in Figure 3 and the optical modulation unit 4200 shown in Figure 4 are used.

[0018] The digital signal processing unit 5100 includes an encoding / mapping unit 5110, a sub-signal generation unit 5120, and a sampling rate f s :f s It includes digital-to-analog converter (DAC) arrays 5161 and 5162 that operate at ≥B. The configuration of the digital signal processing unit 5100 is equivalent to that of a digital signal processing unit in a typical digital coherent optical transmitter, except that the sub-signal generation unit 5120 distributes the transmission symbol sequence to two subchannels.

[0019] The encoding and mapping unit 5110 performs FEC (forward error correction) encoding on the transmitted bit sequence, and then maps the encoded bit sequence to a complex symbol sequence of orthogonal bipolarizations of X and Y. Known training symbols may be inserted into the symbol sequence as needed.

[0020] The sub-signal generation unit 5120 alternately distributes a complex symbol sequence with a symbol rate of 2B to be transmitted for each polarization channel, one symbol at a time, and sets the sampling rate from B to f for each channel. s This is converted to a format, and two sub-signals are generated by performing pulse shaping for bandwidth limiting and pre-equalization using a filter with the inverse characteristics of the device response (f s (In the case of =B, pulse shaping and pre-equalization may not be performed.)

[0021] DAC arrays 5161 and 5162 each output four lanes of analog signals corresponding to the I and Q components of each polarization channel. The output sub-signals are amplified by drivers 5521 and 5522 (four lanes each) to drive the data modulation section of the optical modulation section 5400. The length of the wiring from DAC arrays 5161 and 5162 to the optical modulation section 5400 is adjusted so that the symbol timing matches the timing of the peaks in the intensity waveform of the CSRZ pulse train. However, delay adjustment sections 5511 and 5512 may be provided in the wiring to absorb manufacturing errors, etc. By using such an optical transmitter 5000, a CSRZ-TDM output signal with a symbol rate of 2B can be obtained. [Prior art documents] [Patent Documents]

[0022] [Patent Document 1] Patent No. 6385848 [Non-patent literature]

[0023] [Non-Patent Document 1] H. Yamazaki, A. Sano, M. Nagatani, and Y. Miyamoto, “Single-carrier 1-Tb / s PDM-16QAM transmission using high-speed InP MUX-DACs and an integrated OTDM modulator,” Opt. Express, vol. 23, no. 10, pp. 12866-12873, May. 2015. [Overview of the project]

[0024] As seen in the optical modulator 1000 shown in Figure 1 and the optical modulator 2000 shown in Figure 2, a configuration that uses an optical delay unit to shift the timing of CSRZ pulses by 1 / 4 period between subchannels is the simplest configuration for an optical modulator that generates a CSRZ-TDM signal. However, since it is generally difficult to make the delay τ imparted by the optical delay unit variable, if τ deviates from the ideal value of 1 / (2B) due to changes in the symbol rate or manufacturing errors, such delay errors cannot be compensated for by adjustments in the optical domain. For this reason, it is desirable to be able to compensate for delay errors outside the optical modulator. However, the optical transmitter 5000 shown in Figure 5 cannot compensate for such delay errors. At first glance, it may seem that compensation can be made using the delay adjustment units 5511 and 5512, but in reality, they can only adjust the symbol timing of the drive data signal, and cannot adjust the relative timing between arms of the CSRZ pulse train. Similarly, even if digital delay adjustment units are provided in each path after the sub-signal generation unit 5120 in the digital domain, the above delay errors cannot be compensated for. Therefore, delay errors directly lead to a deterioration of signal quality, and changing the symbol rate requires the preparation of separate optical modulators with different optical delay section designs, which presents a challenge.

[0025] In contrast, the first embodiment in Patent Document 1 and Non-Patent Document 1 disclose an optical modulator that eliminates the optical delay unit and instead uses two Mach-Zehnder modulation circuits for CSRZ pulse generation, performing delay adjustment in the analog electrical domain, thereby enabling flexible adaptation to changes in the symbol rate, etc. However, such optical modulators require two pulse generation units and two clock signals to drive them, resulting in the problem of complex optical circuits and drive systems.

[0026] Furthermore, in all of the above cases, there was a problem in that there was no means to compensate for the effects of spurious emissions (unwanted frequency components) caused by the imperfections in the structure of the pulse generation unit.

[0027] This disclosure has been made in view of these problems, and its purpose is to provide a digital signal processing means that can compensate for delay errors and spurious effects in the digital domain on the transmitting side in a CSRZ (Optical Time-Division Multiplexing: OTDM) transmitter.

[0028] To achieve this objective, one embodiment of the present disclosure is an optical transmitter comprising a digital signal processing unit, an optical CSRZ pulse generation unit, and an optical modulation unit, wherein the optical CSRZ pulse generation unit is configured to output an optical CSRZ pulse train with an electric field waveform frequency of B / 2, and the optical modulation unit comprises an optical branching unit that branches the input optical CSRZ pulse train into two, two data modulation units that data modulate each of the two branched CSRZ pulse trains, an optical coupling unit that couples the output optical signals from the two data modulation units, and an optical delay unit that provides a delay such that the optical CSRZ pulse train passing through one of the two data modulation units is delayed by time τ compared to the optical CSRZ pulse train passing through the other. The digital signal processing unit comprises a sub-signal generation unit that alternately distributes the symbol sequence of the optical signal to be ultimately transmitted from the optical modulation unit one symbol at a time and outputs two sub-signals u1(k) and u2(k): k=0, 1, 2, ...; a first sample shift unit that performs a sample shift on u2(k) corresponding to a delay of 1 / (2B) and outputs u2'(k); a 2x2 calculation unit that obtains v1(k) and v2'(k) from u1(k) and u2'(k) using formula (1) with Δτ=τ-1 / (2B) and t0 as parameters; and a second sample shift unit that performs a sample shift on v2'(k) according to B to obtain v2(k).

[0029] Furthermore, one embodiment of the present disclosure is an optical transmitter comprising a digital signal processing unit, an optical CSRZ pulse generation unit, and an optical modulation unit, wherein the optical CSRZ pulse generation unit is configured to output an optical CSRZ pulse train with an electric field waveform frequency of B / 2, the optical modulation unit comprises an optical branching unit that branches the input optical CSRZ pulse train into two, two data modulation units that data modulate each of the two branched CSRZ pulse trains, an optical coupling unit that couples the output optical signals from the two data modulation units, and an optical delay unit that delays the optical CSRZ pulse train passing through one of the two data modulation units by a time τ relative to the optical CSRZ pulse train passing through the other, and the digital signal processing unit modulates each polarization component of X and Y within approximately ±B frequency. The system comprises a target signal generation unit that virtually generates a target signal consisting of spectral components; a spectral division unit that virtually divides the spectrum of the target signal into four adjacent parts, each having an interval of approximately B / 2, and generates four digital division signals corresponding to signals that have been virtually frequency-shifted to the baseband; an aliasing unit that obtains four aliased digital division signals by inverting each of the four digital division signals on the frequency axis around frequency B / 4 and taking their complex conjugate; and an 8x4 filter unit that receives the four digital division signals and the four aliased digital division signals as inputs and outputs four drive signals corresponding to the I and Q components in the X and Y polarization channels of each of the two data modulation units.

[0030] According to one embodiment of the present disclosure, a digital signal processing means can be provided in a CSRZ-OTDM transmitter that can compensate for delay errors and spurious effects in the digital domain on the transmitting side. [Brief explanation of the drawing]

[0031] [Figure 1] (a) is a diagram of the configuration of a conventional CSRZ-TDM optical modulator, (b) is the electric field waveform of the optical CSRZ pulse train arriving at the output optical port 1002 via arm 1, and (c) is the electric field waveform of the optical CSRZ pulse train arriving at the output optical port 1002 via arm 2. [Figure 2](a) is a diagram of the configuration of a conventional CSRZ-TDM optical modulator, (b) is the electric field waveform of the optical CSRZ pulse train arriving at the output optical port 2002 via arm 1, and (c) is the electric field waveform of the optical CSRZ pulse train arriving at the output optical port 2002 via arm 2. [Figure 3] This is a diagram showing the circuit layout of the optical modulation section. [Figure 4] This is a diagram showing the circuit layout of the optical modulation section. [Figure 5] This is a diagram showing the configuration of a conventional CSRZ-TDM optical modulator. [Figure 6] This is a diagram showing the configuration of a CSRZ-TDM optical modulator according to one embodiment of the present disclosure. [Figure 7] This is a diagram showing the configuration of a CSRZ-TDM optical modulator according to one embodiment of the present disclosure. [Figure 8] This diagram illustrates the operation of the spectral splitting unit, where (a) is the spectrum of the optical signal to be generated and (b) is the output of the spectral splitting unit 7130. [Figure 9] This is a schematic diagram of the spectrum of an optical CSRZ pulse train, where (a) is the spectrum of the CSRZ pulse train output via the first data modulation section of the optical modulation unit 7400, and (b) is the spectrum of the CSRZ pulse train output via the second data modulation section of the optical modulation unit 7400. [Figure 10] This diagram shows the correspondence between the spectrum and symbols of the drive signal, and represents the drive signal of component XI of the first data modulation section of the optical modulation section 7400. [Figure 11] This is an exploded view of the output optical signal spectrum of the optical modulation unit 7400, where (a) is the spectrum of the component modulated by the XI component of the first data modulation unit, (b) is the spectrum of the component modulated by the XQ component of the first data modulation unit, (c) is the spectrum of the component modulated by the XI component of the second data modulation unit, and (d) is the spectrum of the component modulated by the XQ component of the second data modulation unit. [Modes for carrying out the invention]

[0032] The embodiments of this disclosure will be described in detail below with reference to the drawings. Note that the following description is illustrative, and embodiments with some configurations modified are possible without departing from the gist of this disclosure.

[0033] (First embodiment) Figure 6 shows the configuration of an optical transmitter 6000, which is a first embodiment of the present disclosure. The optical transmitter 6000 comprises a digital signal processing unit 6100, a transmitting laser 6200, an optical CSRZ pulse generation unit 6300, an optical modulation unit 6400, delay adjustment units 6511 and 6512, and drivers 6521 and 6522.

[0034] The optical CSRZ pulse generator 6300 corresponds to the frequency B / 2 clock source 1151 and pulse generator 1111 in the optical modulator 1000 shown in Figure 1, or the frequency B / 2 clock source 2151 and pulse generator 2111 in the optical modulator 2000 shown in Figure 2.

[0035] The optical modulation section 6400 corresponds to the portion of the optical modulator 1000 shown in Figure 1 from the optical branching section 1231 to the optical multiplexing section 1232, or to the portion of the optical modulator 2000 shown in Figure 2 from the optical branching section 2231 to the optical multiplexing section 2232. Let τ be the delay provided by the optical delay section corresponding to the optical delay section 1241 in Figure 1 or the optical delay section 2241 in Figure 2. The data modulation sections 1221 and 1222 in Figure 1 or the data modulation sections 2221 and 2222 in Figure 2 use polarization-multiplexed optical IQ modulation sections laid out as the optical modulation section 3200 shown in Figure 3 and the optical modulation section 4200 shown in Figure 4.

[0036] The digital signal processing unit 6100 includes an encoding / mapping unit 6110, a sub-signal generation unit 6120, sample shift units 6131, 6132, and 6133, a 2x2 calculation unit 6140, and a sampling rate f s :f s It includes digital-to-analog converter (DAC) arrays 6161 and 6162 that operate at ≥B.

[0037] The encoding / mapping unit 6110 performs FEC encoding on the transmit bit sequence, and then maps the encoded bit sequence onto complex symbol sequences of two orthogonal polarizations X and Y. Known training symbols may be inserted into the symbol sequences as necessary. Since subsequent digital signal processing is basically performed for each polarization channel, the following description focuses only on the X polarization channel for simplicity, but similar processing is also performed for the Y polarization channel.

[0038] The sub-signal generation unit 6120 alternately distributes the complex symbol sequence with a symbol rate of 2B to be transmitted for each symbol, and converts the sampling rate from B to f s , and performs pulse shaping for band limitation, pre-equalization using a filter with the inverse characteristic of the device response, and the like, to generate two systems of complex sub-signals (f s =B, pulse shaping or pre-equalization may not be performed in some cases). Specifically, for each polarization channel, the sub-signal generation unit 6120, assuming that the complex symbol sequence to be transmitted is z(n):n=0, 1, 2, ..., alternately distributes z(n) to generate a first sub-symbol sequence z1(m)=(-1) m z(2m):m=0, 1, 2, ... and a second sub-symbol sequence z2(m)=(-1) m z(2m+1):m=0, 1, 2, .... Furthermore, the sub-signal generation unit 6120 performs sampling rate conversion, pulse shaping, pre-equalization, and the like on z1(m) and z2(m) respectively, and outputs sub-signals u1(k) and u2(k):k=0, 1, 2, ..., where f s =B, u1(k)=z1(k) and u2(k)=z2(k) may be simply set in some cases.

[0039] The sample shift unit 6131 performs a sample timing shift corresponding to a time delay of 1 / (2B) on u2(k), and outputs u2´(k).

[0040] The 2×2 operation unit 6140 outputs v1´(k) and v2´(k) by the operation represented by Equation (1). TIFF0007939594000001.tif15150

[0041] However, Δτ = τ - 1 / (2B), and t0 is a parameter relating to the relative timing difference between the optical CSRZ pulse train and the drive signal, as described later.

[0042] The sample shift units 6133 and 6132 shift the sample timing of v1'(k) and v2'(k), respectively, and output v1(k) and v2(k). The amount of shift here depends on B and the wiring length, as described later.

[0043] DAC arrays 6161 and 6162 each output four lanes of analog signals corresponding to the I and Q components of the orthogonal two-polarization channels. The output sub-signals are amplified by drivers 6521 and 6522 (four lanes each) to drive the data modulation section of the optical modulation section 6400. The length of the wiring from DAC arrays 6161 and 6162 to the optical modulation section 6400 is adjusted so that the symbol timing matches the timing of the peaks in the intensity waveform of the CSRZ pulse train. However, delay adjustment sections 6511 and 6512 may be provided in the wiring to absorb manufacturing errors, etc. By using such an optical transmitter 6000, a CSRZ-TDM output signal with a symbol rate of 2B can be obtained.

[0044] The following analytical explanation describes the principle of delay error compensation by the sample shift units 6131, 6132, 6133 and the 2x2 calculation unit 6140. For simplicity, only X-polarized signals are considered in the following explanation, but the same method of compensation is applied to Y-polarized signals. Furthermore, when expressing the electric field waveform mathematically, the carrier component of the transmitted laser frequency is omitted.

[0045] First, the electric field waveform of the CSRZ pulse in this example can be expressed by equation (2), if we ignore higher-order terms for simplicity. Here, t0 is a parameter corresponding to the time origin of the pulse electric field waveform, and can be adjusted inside the optical CSRZ pulse generator 6300 using a phase shifter or the like. TIFF0007939594000002.tif6150

[0046] Let s1(t) be the waveform of the complex data signal sent to the first data modulation section within the optical modulation section 6400, and s2(t) be the waveform of the complex data signal sent to the second data modulation section within the optical modulation section 6400. Furthermore, let x be the electric field waveform of the final optical output signal. out (t), of which the component passing through the first data modulation section is x1(t), and the component passing through the second data modulation section is x2(t). For simplicity, if we ignore the modulation efficiency and the scale factor representing optical loss, x out (t), x1(t), and x2(t) can be expressed as equations (3a), (3b), and (3c), respectively. TIFF0007939594000003.tif20150

[0047] However, t1 and t2 are time shifts of the drive signals s1 and s2, respectively, and are determined by the wiring design, but can also be adjusted using the delay adjustment units 6511 and 6512.

[0048] Ideally, τ = 1 / (2B), but due to changes in the symbol rate, manufacturing tolerances, etc., τ may not equal 1 / (2B). In this case, simply adjusting t0, t1, and t2 will not compensate for the deviation in τ, i.e., the delay error. This is because, as is clear from equations (3b) and (3c), the timing difference between s1 and s2 between subchannels can be arbitrarily set by adjusting t1 and t2, but the electric field waveform x of the CSRZ pulse between subchannels... CSRZ The timing difference is determined solely by τ. Thus, in a CSRZ-TDM optical transmitter, the delay error provided by the optical delay section cannot be compensated for by intuitive methods such as adjusting the timing of the clock and each sub-signal.

[0049] The principle of delay error compensation in this disclosure is based on the idea of ​​compensating for delay errors by performing 2-input, 2-output arithmetic processing in the digital domain, rather than individually adjusting the delay of the sub-signals that have been distributed in the sub-signal generation unit. To proceed with the explanation, first consider the output waveform x in the ideal case with no error, i.e., τ = 1 / (2B). out、ideal Consider (t). In this case, the 2x2 arithmetic unit 6140 simply needs to set v1(k)=u1(k) and v2(k)=u2(k), and also t0=t1=t2=0. Let s1(t) and s2(t) be the cases where τ=1 / (2B) and v1(k)=u1(k) and v2(k)=u2(k), respectively. 1、ideal (t) and s 2、ideal If we represent it as (t), then the output waveform x out、ideal (t) can be expressed as equation (4). TIFF0007939594000004.tif9150

[0050] The first term on the right-hand side of equation (4) is zero for t=2m / 2B: m=0, 1, 2, ..., and the second term is zero for t=(2m+1) / 2B: m=0, 1, 2, .... If h1(t) and h2(t) are the analog response from DAC arrays 6161 and 6162 to the optical modulation section 6400 and the impulse response including pulse shaping and pre-equalization within the sub-signal generation section 6120, then s 1、ideal (t) and s 2、ideal (t) can be expressed as equation (5a) and equation (5b), respectively. TIFF0007939594000005.tif26150

[0051] To simplify the explanation, let's first consider the case where there is no intersymbol interference at the output of the optical modulation unit 6400, that is, when h1(t) and h2(t) each take a value of 1 at t=0 and a value of 0 at t=±m / B:m=1, 2, 3, ... (when pre-equalization is performed by the sub-signal generation unit 6120 to achieve this), and then the output waveform x out、ideal (t) becomes equation (6a), equation (6b), and equation (6c), and indeed x out、idealIt can be confirmed that (t) is a signal with a symbol rate of 2B for symbol values ​​z(n): n=0, 1, 2, ... TIFF0007939594000006.tif39150

[0052] Next, consider the general case with errors, i.e., the case where τ ≠ 1 / (2B). In this case, the electric field waveform x of the final optical output signal is... out (t) can be expressed by equation (3a). x expressed by equation (3a) out (t) is x represented by formula (4) out、ideal If we can adjust s1(t) and s2(t) so that they are equal to (t), then we can say that we have compensated for the error in the group delay difference. In other words, we just need to determine s1(t) and s2(t) so that equation (6) holds. TIFF0007939594000007.tif15150

[0053] Let τ = 1 / (2B) + Δτ, and expanding equation (6) yields equation (7). TIFF0007939594000008.tif70150

[0054] Comparing the coefficients of cos(πBt) and sin(πBt) in equation (7), we can see that if equation (8) is true, then equation (6) is true. TIFF0007939594000009.tif39150

[0055] Multiplying both sides of equation (8) by the inverse of the 2x2 matrix on the right side from the left and rearranging gives equation (9). TIFF0007939594000010.tif39150

[0056] The sample shift units 6131, 6132, 6133 and the 2x2 calculation unit 6140 in this disclosure perform an operation corresponding to formula (9) in the digital domain. In order to perform this process properly, s 1、ideal (t) and s 2、ideal The spectral component of (t) is ±f, which is the Nyquist frequency. sIt is desirable that it stays well within / 2. If the DAC's analog bandwidth is sufficiently smaller than the Nyquist frequency, this condition is naturally met, but if the DAC's analog bandwidth is larger than the Nyquist frequency, it is desirable to introduce an anti-aliasing filter in the analog domain on the DAC output side. In the following, s 1、ideal (t) and s 2、ideal Among the spectral components of (t), f > f s / 2 and f < f s The proportion of the / 2 component is sufficiently small, and aliasing can be considered negligible. Then, let u1(k) and u2(k) be s 1、ideal (t) and s 2、ideal (t) The sampling rate f s Using the discrete signal sampled, the processing equivalent to equation (8) can be performed using equation (1) above. The shift amounts in sample shift units 6133 and 6132 correspond to delays of -t1 and -t2-Δτ-1 / (2B), respectively, based on equation (8). If t1 can be set to 0 by wiring design or adjustment of delay adjustment unit 6511, sample shift unit 6133 becomes unnecessary, but sample shift unit 6132 is necessary to ensure flexibility in changing the symbol rate B and cannot be omitted.

[0057] In this way, by using the sample shift units 6131, 6132, 6133 and the 2x2 calculation unit 6140 within the digital signal processing unit 6100, even when there is a delay error Δτ due to symbol rate changes or manufacturing errors, the final optical output signal can be equivalent to that of a signal where Δτ does not exist.

[0058] It is desirable to measure the value of τ during the factory inspection of the optical modulation unit 6400, but it can also be measured using a known test signal after it has been incorporated into the optical transmitter 6000.

[0059] (Second embodiment) Figure 7 shows the configuration of an optical transmitter 7000, which is a second embodiment of the present disclosure. The optical transmitter 7000 comprises a digital signal processing unit 7100, a transmitting laser 7200, an optical CSRZ pulse generation unit 7300, an optical modulation unit 7400, delay adjustment units 7511 and 7512, and drivers 7521 and 7522.

[0060] The optical CSRZ pulse generator 7300 corresponds to the frequency B / 2 clock source 1151 and pulse generator 1111 in the optical modulator 1000 shown in Figure 1, or the frequency B / 2 clock source 2151 and pulse generator 2111 in the optical modulator 2000 shown in Figure 2.

[0061] The optical modulation section 7400 corresponds to the portion of the optical modulator 1000 shown in Figure 1 from the optical branching section 1231 to the optical multiplexing section 1232, or to the portion of the optical modulator 2000 shown in Figure 2 from the optical branching section 2231 to the optical multiplexing section 2232. Let τ be the delay provided by the optical delay section corresponding to the optical delay section 1241 in Figure 1 or the optical delay section 2241 in Figure 2. The data modulation sections 1221 and 1222 in Figure 1 or the data modulation sections 2221 and 2222 in Figure 2 use polarization-multiplexed optical IQ modulation sections laid out as the optical modulation section 3200 shown in Figure 3 and the optical modulation section 4200 shown in Figure 4.

[0062] The digital signal processing unit 7100 includes an encoding / mapping unit 7110, a target signal generation unit 7120, a spectral division unit 7130, an aliasing unit 7140, an 8x4 filter unit 7150, and a sampling rate f s :f s It includes digital-to-analog converter (DAC) arrays 7161 and 7162 that operate at ≥B.

[0063] The encoding and mapping unit 7110 performs FEC encoding on the transmitted bit sequence and then maps the encoded bit sequence to a complex symbol sequence of X and Y orthogonal dual polarizations. Known training symbols may be inserted into the symbol sequence as needed. In this embodiment, the complex symbol sequence does not necessarily have to be a single-carrier system with one sequence per polarization channel; a multi-carrier system in which multiple symbol sequences are allocated to multiple RF subcarriers is also possible. The symbol rate can also be any value of 2B or less. Subsequent digital signal processing is basically performed for each polarization channel, so for simplicity, we will explain focusing only on the X-polarization channel, but the same processing is performed for the Y-polarization channel.

[0064] The target signal generation unit 7120 virtually generates the signal waveform that is to be generated around the optical carrier as the output of the optical transmitter 7000, depending on the input complex symbol sequence. For example, if the target signal is a single-carrier signal, the desired pulse waveform is simply convolved into the single sequence of symbols received from the coding and mapping unit 7110. If the target signal is a multi-carrier signal, the desired pulse waveform is convolved into each of the multiple sequences of symbols received from the coding and mapping unit 7110, and then added together. However, the spectral components of the target signal must be concentrated within approximately ±B. Therefore, the symbol rate in the case of a single-carrier signal must basically be 2B or less.

[0065] Figure 8 illustrates the operation of the spectral splitting unit 7130. The spectrum of the X-polarization component of the target signal sent from the target signal generation unit 7120 is divided into U Xtot (f) The optical carrier frequency is f c In this case, the spectrum of the optical signal to be generated is, as shown in Figure 8(a), U in complex notation. Xtot (ff c ) can be expressed as follows. The spectral splitting section 7130 is U as shown in Figure 8(b). Xtot (ff c ) to f cThe spectrum U is virtually divided into four parts approximately every B / 2 in width around it, and these are then virtually shifted to the baseband. X1 (f), U X2 (f), U X3 (f), U X4 It outputs four digital division signals represented by (f). As shown in Figure 8, U Xtot (ff c )=U X1 (f+2B-f c )+U X2 (f+Bf c )+U X3 (ff c )+U X4 (fBf c )

[0066] The folded portion 7140 is U X1 (f), U X2 (f), U X3 (f), U X4 Each of (f) is inverted around B / 4 on the frequency axis, and its complex conjugate is taken to output four aliased digital division signals. The spectra of the four aliased digital division signals are expressed as follows, where a superscript asterisk represents the complex conjugate. TIFF0007939594000011.tif13150

[0067] The 8x4 filter unit 7150 takes four digital division signals and four aliased digital division signals as inputs and outputs four signals through 8-input, 4-filter processing. The method for determining the filter coefficients of the 8x4 filter unit 7150 will be described later. The four output signals of the 8x4 filter unit 7150 are real signals corresponding to the I and Q components of the first and second sub-signals of the polarization, respectively. Here, the first sub-signal is the signal that drives the first data modulation section of the optical modulation unit 7400 via the DAC array 7161 and driver 7521. The second sub-signal is the signal that drives the second data modulation section of the optical modulation unit 7400 via the DAC array 7162 and driver 7522. The drivers 7521 and 7522 of the DAC arrays 7161 and 7162 and beyond are the same as the drivers 6521 and 6522 of the DAC arrays 6161 and 6162 and beyond in the first embodiment described above, so their explanation will be omitted.

[0068] In this embodiment, not only is the delay error compensated for using the compensation method described in the first embodiment compensated for, but the spurious effects caused by the imperfections of the optical CSRZ pulse generation unit 7400 can also be compensated. Furthermore, it is not limited to single-carrier signals with a symbol rate of 2B, but any signal can be generated as long as the spectral components are concentrated within ±B of the optical carrier frequency. The principle is explained below.

[0069] Figure 9 schematically shows the spectrum of an optical CSRZ pulse train when there is an imperfection in the optical CSRZ pulse generation unit 7400. Figure 9(a) shows the spectrum of the CSRZ pulse train output via the first data modulation unit of the modulation unit 7400, and Figure 9(b) shows the spectrum of the CSRZ pulse train output via the second data modulation unit of the modulation unit 7400. The main component is f c In addition to the ±B / 2 component, f c Spurious signals are generated around the frequency at integer multiples of B / 2. For example, when a null-biased push-pull driven Mach-Zehnder modulator is used as the optical CSRZ pulse generator and driven with a clock of frequency B / 2, spurious signals are generated around the frequency at f due to the nonlinearity (sine wave response) of the output optical field with respect to the drive voltage. c±3B / 2, f c Components of ±5B / 2, ... are generated, and furthermore, due to optical loss between the arms of the Mach-Zehnder modulator, phase modulation efficiency, or imbalance in the drive voltage amplitude, f c ,f c ±B, f c Components of ±2B, ... are generated. As shown in Figure 9(a), the complex electric field amplitudes of each component of the spectrum of the optical CSRZ pulse train output via the first data modulation section of the optical modulation section 7400 are a0, a ±1 a ±2 ...and so, the complex electric field amplitudes of each component of the optical CSRZ pulse train output via the second data modulation unit in the optical modulation unit 7400 are a0, ±1 exp(±jπBτ), a ±2 exp(±j2πBτ), ...

[0070] Now, when such an optical CSRZ pulse train containing spurious signals is input to the optical modulation unit 7400, and modulated and combined by sub-signals sent from DACs 7161 and 7162, the spectrum of the drive signal is convolved into each spurious signal, including the main component mentioned above, in the frequency domain, and an optical signal with a spectrum in which each component overlaps in a complex way is output from the optical modulation unit 7400.

[0071] Here, we consider dividing the spectrum of each drive signal of the optical modulation unit 7400 into two parts, a positive frequency side and a negative frequency side, as shown in Figure 10. Since it is a real signal, the component obtained by shifting the negative frequency side component by +B / 2 to the positive frequency side is equal to the component obtained by folding the positive frequency side component around B / 4 and taking its complex conjugate. For example, if the positive frequency side component of the spectrum of the drive signal of component XI of the first data modulation unit of the optical modulation unit 7400 is X I1 If we denote this as (f), the component obtained by shifting the negative frequency component to the positive frequency side by +B / 2 can be expressed using the tilde symbol mentioned above as follows: TIFF0007939594000012.tif6150

[0072] Similarly, X Q1 (f), X I2 (f), XQ2 For (f), the components obtained by shifting the negative frequency component to the positive frequency side by +B / 2 can be expressed as follows using the tilde symbol mentioned above. TIFF0007939594000013.tif6150

[0073] Based on the above, Figure 11 schematically shows the state in which each of the drive signals of the optical modulation unit 7400 is convolved into the spectrum of the optical CSRZ pulse as shown in Figure 9. Figure 11(a) shows the spectrum of the component modulated by the XI component of the first data modulation unit modulated the optical CSRZ pulse train, Figure 11(b) shows the spectrum of the component modulated by the XQ component of the first data modulation unit modulated the optical CSRZ pulse train, Figure 11(c) shows the spectrum of the component modulated by the XI component of the second data modulation unit modulated the optical CSRZ pulse train, and Figure 11(d) shows the spectrum of the component modulated by the XQ component of the second data modulation unit modulated the optical CSRZ pulse train. To avoid complexity in the figures, only some symbols are shown. The superposition of all of these results in the final optical output signal of the optical transmitter 7000. Comparing Figure 11 and Figure 8, the target signal U is the final optical output signal of the optical transmitter 7000. tot (ff c In order to obtain ), it is sufficient that the relationships shown in the following equations (10a) and (10b) hold. TIFF0007939594000014.tif83153

[0074] The coefficients of the 8×4 filter section 7150 must be determined such that equations (10a) and (10b) are satisfied. Of the four output signals of the 8×4 filter section 7150, the positive frequency components of the spectra of the signals sent to the XI and XQ modulation sections of the first data modulation section of the optical modulation section 7400 are V, respectively. XI1 (f) and V XQ1 (f) The positive frequency components of the spectrum of the signals sent to the XI and XQ modulation sections of the second data modulation section of the optical modulation section 7400 are V, respectively. XI2 (f) and V XQ2 (f) Then X I1 (f), X Q1(f), X I2 (f), X Q2 (f) are each multiplied by the frequency response of the path from DAC 7161 and 7162 to the optical modulation section 7400, so they can be expressed as in Equation (11). XI1 (f), V XQ1 (f), V XI2 (f), V XQ2 (f) are each multiplied by the frequency response of the path from DAC 7161 and 7162 to the optical modulation section 7400, so they can be expressed as in Equation (11). TIFF0007939594000015.tif22150

[0075] T(f) is basically a 4×4 matrix with all elements other than the diagonal elements being zero, but when crosstalk between components exists inside the drivers 7521 and 7522 or the optical modulation section 7400, the elements other than the diagonal elements also become non-zero. In any case, from Equation (10a), Equation (10b) and Equation (11), the spectrum U X1 (f), U X2 (f), U X3 (f), U X4 (f) can be expressed as in Equation (12). TIFF0007939594000016.tif64150

[0076] H 4×8 (f) basically includes not only the frequency response of the path from DAC 7161 and 7162 to the optical modulation section 7400, but also all effects of spurious components and delay errors as shown in FIG. 9. To obtain the matrix for calculating V from U, that is, the compensation matrix, it is sufficient to obtain a matrix equivalent to the inverse matrix of H 4×8 (f). For that purpose, first H 4×8 (f) is decomposed into two 4×4 matrices H L (f) and H R (f), expressed as H 4×8 (f)=(H L (f), H R (f). Then, by expanding Equation (12) in a form that adds the aliasing signal to the U vector, it can be expressed as Equation (13). TIFF0007939594000017.tif43150

[0077] Therefore, the vector V on the right-hand side of equation (13) is given by equations (14a) and (14b). G 8×8 (f) Two 4x8 matrices G U (f) and G L If we decompose it into (f) and express it as in equation (15), we get equation (16). TIFF0007939594000018.tif127150

[0078] This G U (f) is the coefficient of the 8×4 filter section 7130. More precisely, the derived V XI1 (f), V XQ1 (f), V XI2 (f), V XQ2 (f) is a spectrum containing only positive frequency components, so the function of the 8x4 filter section 7130 is to perform an inverse Fourier transform on these components and then send the resulting signal (after removing the real part) to the DAC7161 and 7162.

[0079] To actually determine the coefficients of the 8x4 filter section 7130, first use known test signals etc. 4×8 (f) was measured, and based on that, G was derived from equations (14b) and (15). U You can find (f) and use it as the coefficient for the 8×4 filter section 7130.

[0080] It should be noted that the explanation so far assumes that aliasing does not occur in DAC7161 and 7162. For example, if the tail of the drive signal spectrum extends beyond ±B / 2 as shown in Figure 10, the tail of the spectrum becomes the Nyquist band, i.e., ±f s We have been considering the assumption that it is well within / 2. The analog response bandwidth of DAC7161 and 7162 is f s If the value is sufficiently smaller than / 2, aliasing will not occur, but otherwise, it is desirable to prevent aliasing using an anti-aliasing filter or the like. Alternatively, f s Setting it to =B eliminates the problem even if aliases exist. This is because in that case, ±B / 2 = ±f s Aliases appearing outside of / 2 are the main signals inside ±B / 2 = ±fs Since it is a folded signal that has been folded back by / 2, in the model represented by equations (10) and later, the alias can be completely incorporated as the signal component of the adjacent frequency slot, and G in equation (16) U This is because the effect of aliasing can also be compensated for by the 8×4 filter section 7150 with coefficient (f). s Setting =B means f s Since the ratio of the frequency B / 2 of the optical CSRZ pulse is a simple integer ratio of 2:1, there is also the advantage that clock synchronization becomes easier in terms of implementation.

[0081] As explained above, by using an optical transmitter 7000 that includes a target signal generation unit 7120, a spectral division unit 7130, an aliasing unit 7140, and an 8x4 filter unit 7150 in the digital signal processing unit 7100, it is possible to compensate not only for the delay error of the optical delay unit in the optical modulation unit 7400 but also for the spurious effects caused by the imperfections of the optical CSRZ pulse generation unit 7300. Furthermore, it is possible to generate any signal within a signal bandwidth of ±B of the optical carrier frequency, not limited to a single-carrier signal with a symbol rate of 2B. [Industrial applicability]

[0082] According to one embodiment of this disclosure, a CSRZ-OTDM transmitter can be provided that can compensate for delay errors and spurious effects in the digital domain of the transmitting side. [Explanation of Symbols]

[0083] 1000, 2000 optical modulators Optical ports for inputs 1001 and 2001 1002, 2002 Optical output ports 1100, 2100 Optical CSRZ pulse generation unit 1111, 2111 pulse generation unit 1151, 2151 clock sources 1200, 2200, 3200, 4200 Optical Modulation Section 1221, 2221, First data modulation section 1222, 2222 Second data modulation section 1231, 2231, 3231, 3261, 3262, 4231, 4261, 4262 Optical branching section 1232, 2232 Optical multiplexing section 1241, 2241, 3241 Optical delay section 3221, 4221 First polarization multiplexed IQ modulation section 3221-X, 3221-Y, 3222-X, 3222-Y IQ Modulation Unit 3270, 3270, 4270, 4270 Polarization 90-degree rotation section 3280, 3280, 4280 polarization coupling section 3222, 4222 Second polarization multiplexed IQ modulation section 3232-X, 4232-XX Polarization side optical multiplexer 3232-Y, 4232-YY polarization side optical multiplexer 4241-XX polarization side optical delay section 4241-YY polarization side optical delay section 4221-X, 4222-XX polarization-side IQ modulation section 4221-Y, 4222-YY polarization-side IQ modulation section 5000, 6000, 7000 Optical Transmitters 5100, 6100, 7100 Digital Signal Processing Unit 5110, 6110, 7110 Encoding and Mapping Section 5120, 6120 Sub-signal generation unit 5161, 5162, 6161, 6162, 7161, 7162 Digital-to-Analog Converter (DAC) Arrays 5200, 6200, 7200 transmitting laser 5300, 6300, 7300 Optical CSRZ pulse generation unit 5400, 6400, 7400 Optical Modulation Section 5511, 5512, 6511, 6512, 7511, 7512 Delay adjustment section Drivers 5521, 5522, 6521, 6522, 7521, and 7522 6131, 6132, 6133 Sample shift section 6140 2×2 operation section 7120 Target signal generation unit 7130 Spectral division section 7140 Folded section 7150 8x4 filter section

Claims

1. An optical transmitter comprising a digital signal processing unit, an optical CSRZ pulse generation unit, and an optical modulation unit, The optical CSRZ pulse generation unit is configured to output an optical CSRZ pulse train with an electric field waveform frequency of B / 2. The optical modulation unit is An optical branching unit that splits the input optical CSRZ pulse train into two, Two data modulation units that data modulate each of the two-branched CSRZ pulse trains, An optical coupling unit that combines the output optical signals from the two data modulation units, An optical delay unit that provides a delay such that an optical CSRZ pulse train passing through one of the two data modulation units is delayed by time τ relative to an optical CSRZ pulse train passing through the other unit. Equipped with, The aforementioned digital signal processing unit is The optical modulation unit then alternately distributes the symbol sequence of the optical signal to be ultimately transmitted, one symbol at a time, to two sub-signals u 1 (k) and u 2 (k): A sub-signal generation unit that outputs k = 0, 1, 2, ... u 2 A sample shift equivalent to a delay of 1 / (2B) is performed on (k) u 2 The first sample shift section outputs '(k), u 1 (k) and u 2 '(k), Δτ=τ-1 / (2B), t 0 as a parameter to obtain v 1 (k) and v 2 '(k), a 2×2 operation unit, v 2 Perform a sample shift corresponding to B for '(k) v 2 A second sample shift unit that obtains (k) and An optical transmitter equipped with [a specific feature].

2. An optical transmitter comprising a digital signal processing unit, an optical CSRZ pulse generation unit, and an optical modulation unit, The optical CSRZ pulse generation unit is configured to output an optical CSRZ pulse train with an electric field waveform frequency of B / 2. The optical modulation unit is An optical branching unit that splits the input optical CSRZ pulse train into two, Two data modulation units that data modulate each of the two-branched CSRZ pulse trains, An optical coupling unit that combines the output optical signals from the two data modulation units, An optical delay unit that provides a delay such that an optical CSRZ pulse train passing through one of the two data modulation units is delayed by time τ relative to an optical CSRZ pulse train passing through the other unit. Equipped with, The aforementioned digital signal processing unit processes each of the X and Y polarization components as follows: A target signal generation unit virtually generates a target signal consisting of spectral components within approximately ±B of frequency, A spectrum division unit virtually divides the spectrum of the target signal into four adjacent parts, each having an interval of approximately B / 2, and generates four digital division signals that correspond to signals that have been virtually frequency-shifted to the baseband. The folding section obtains four aliased digital divided signals by inverting each of the four aforementioned digital divided signals around frequency B / 4 on the frequency axis and taking their complex conjugate, An 8x4 filter unit that receives the four digital division signals and the four aliased digital division signals and outputs four drive signals corresponding to the I and Q components in the X and Y polarization channels of each of the two data modulation units. An optical transmitter characterized by having the following features.

Images