JP2006171677A - Optical pulse compression device, optical pulse generating device and optical pulse compression method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To optimize the dispersion profile of a fiber which compresses optical pulses, and to generate high-quality compressed pulses. <P>SOLUTION: The optical pulse generation device 100 is composed of a beat light generating unit 1 generating beat light and an optical pulse compressing device 2, where the optical pulse compressing device 2 is composed of a converting unit 3 and a compressing unit 4. The conversion unit 3 is composed of a DDF (dispersion reduced fiber) in which an abnormal dispersion value decreases in the longitudinal direction from the input end to the output end and beat light inputted from the beat light generating unit 1 is converted into up-chirp pulses. The compressing unit 4 is composed of a DDF, in which the abnormal dispersion value reduces proportionally to the position in the longitudinal direction, and time width of the up-chirp pulses inputted from the converting unit 3 is compressed to output an optical pulse train. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an optical pulse compression device that compresses an optical pulse, an optical pulse generation device including the optical pulse compression device, and an optical pulse compression method.

  Conventionally, an optical pulse compression device using an optical fiber has been widely used. The optical fiber is characterized by its high-efficiency nonlinear medium due to its low loss and excellent dispersion controllability, and various optical pulse compression devices that utilize these features are proposed. Has been.

  Among these, as shown in Non-Patent Document 1, a dispersion decreasing fiber (DDF: Dispersion Decreasing Fiber) having anomalous dispersion and gradually reducing the absolute value of the dispersion value in the longitudinal direction is a comparison. It is known as a means capable of generating a high-quality optical pulse.

In addition, as shown in Patent Document 1, a comb-like profiled fiber (CPF) in which a highly nonlinear fiber and a highly dispersive fiber are alternately connected with an appropriate length is also available. Similar to DDF, it is known as a means capable of generating a relatively high quality optical pulse. Furthermore, as shown in Patent Document 2, it is known that DDF can be used to generate ultra-wideband light.
JP 2004-117590 A Japanese Patent Laid-Open No. 10-90737 SV Chernikov, EM Dianov, DJ Richardson and DN Payne, “Soliton pulse compression in dispersion-decreasing fiber,” Opt. Lett., Vol. 18, no. 7, pp. 476-478, 1993.

  As described above, it is known that an optical pulse compression apparatus using DDF and CPF can generate high-quality optical pulses in a comparator. However, these DDFs and CPFs require the design of dispersion values and lengths that match the parameters of the input pulse and output pulse. However, when the longitudinal distribution of these dispersion values is not optimal, the quality of the optical pulse is reduced. There was a problem of doing. For example, when the variation of the dispersion value is abrupt, an uncompressed component called a pedestal accompanies the compressed light pulse. On the other hand, when the dispersion value fluctuates gently, a longer fiber is necessary, so that the dispersion effect accumulates and increases timing jitter and the like.

  An object of the present invention is to optimize a dispersion profile (dispersion distribution) of a fiber for compressing an optical pulse and to generate a high-quality compressed pulse.

  In order to solve the above-mentioned problem, the invention described in claim 1 is directed to an optical pulse compression apparatus that inputs beat light and compresses the time width of the beat light to generate an optical pulse. The anomalous dispersion value decreases, and the anomalous dispersion value decreases in proportion to the position in the direction from the input end to the output end, and the time width of the up chirp pulse decreases. And a compressing unit for compressing.

  According to a second aspect of the present invention, in the optical pulse compression device according to the first aspect of the invention, in the compression unit, the anomalous dispersion value is halved at a distance of 4 to 6 times the dispersion length at the input end. Yes.

  According to a third aspect of the present invention, in the optical pulse compression device according to the first or second aspect, the conversion unit and / or the compression unit is a dispersion reducing fiber.

  According to a fourth aspect of the present invention, in the optical pulse compression device according to the first or second aspect, the conversion unit and / or the compression unit is a comb-shaped profile fiber in which nonlinear fibers and linear fibers are alternately connected. It is characterized by that.

  According to a fifth aspect of the present invention, in the optical pulse compression device according to the fourth aspect, the comb-shaped profile fiber is disposed next to the longitudinal length of the nonlinear fiber disposed at the input end. It is characterized by being shorter than a nonlinear fiber.

  The invention described in claim 6 is characterized by comprising the beat light generating section for generating beat light and the optical pulse compression device according to any one of claims 1 to 5.

  The invention according to claim 7 generates beat light, converts the beat light into an up-chirp pulse using a converter whose anomalous dispersion value decreases in the direction from the input end to the output end, and from the input end to the output end. The time width of the up chirp pulse is compressed using a compressor whose anomalous dispersion value decreases in proportion to the position in the direction of.

  The invention according to claim 8 is the optical pulse compression method according to claim 7, wherein the up chirp is obtained by using a compressor whose anomalous dispersion value is halved at a distance of 4 to 6 times the dispersion length at the input end. It is characterized by compressing the time width of the pulse.

  A ninth aspect of the present invention is the optical pulse compression method according to the seventh or eighth aspect, wherein the converter and / or the compressor is a dispersion reducing fiber.

  A tenth aspect of the present invention is the optical pulse compression method according to the seventh or eighth aspect, wherein the converter and / or the compressor is a comb profile fiber in which nonlinear fibers and linear fibers are alternately connected. It is characterized by that.

  According to an eleventh aspect of the present invention, in the optical pulse compression method according to the tenth aspect, the comb-shaped profile fiber is disposed next to the longitudinal length of the nonlinear fiber disposed at the input end. It is characterized by being shorter than a nonlinear fiber.

  According to the present invention, beat light is converted into an up-chirp pulse, and the time width of the up-chirp pulse is compressed, so that the generation of a pedestal can be suppressed and a compressed pulse with less jitter can be obtained. The quality of the light pulse can be improved.

  In addition, by using beat light as seed light, an ultrahigh frequency signal can be easily created, and optical pulse generation utilizing the ultrahigh frequency signal can be realized.

  Hereinafter, first to third embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
A first embodiment of the present invention will be described with reference to FIGS.
First, the configuration in the first embodiment will be described. FIG. 1 shows a configuration of an optical pulse generator 100 according to the first embodiment. As shown in FIG. 1, the optical pulse generator 100 includes a beat light generator 1 and an optical pulse compressor 2. The optical pulse compression device 2 includes a conversion unit (converter) 2 and a compression unit (compressor) 4.

  The beat light generator 1 generates beat light whose intensity changes at the difference frequency between the two laser lights by superimposing two laser lights having different frequencies, and amplifies the generated beat light by an optical amplifier (not shown). And output to the converter 3 of the optical pulse compressor 2. The advantage of using beat light is that it can easily generate ultra-high frequency signals of 100 GHz or higher. For example, when beat light is generated using two semiconductor lasers, the frequency can be freely controlled by appropriately controlling the wavelength of the laser light via the injection current and temperature to these semiconductor lasers. Can do.

  The conversion unit 3 is composed of DDF (Dispersion Decreasing Fiber) whose anomalous dispersion value decreases in the longitudinal direction from the input end to the output end, and converts the beat light input from the beat light generation unit 1 into an up chirp pulse. The up chirp pulse is output to the compression unit 4.

  The compression unit 4 is configured by a DDF (see FIG. 2) in which the anomalous dispersion value decreases in proportion to the position (distance) z in the longitudinal direction, compresses the time width of the up chirp pulse input from the conversion unit 3, Output as an optical pulse train.

  Hereinafter, the dispersion profile of the compression unit 4 and the dispersion profile of the conversion unit 3 will be described in detail.

式（１）において、Ａ(z,t)は光パルスの複素電界であり、β₂及びγは、それぞれ、光ファイバの分散値、非線形定数である。 <Distributed profile of compression section>
First, a general interpretation of soliton compression in DDF will be described. It is known that the change in waveform associated with propagation of an optical pulse through an optical fiber can be well described by the nonlinear Schroedinger equation (1).

In equation (1), A (z, t) is the complex electric field of the optical pulse, and β ₂ and γ are the dispersion value and nonlinear constant of the optical fiber, respectively.

式（２）において、Ｐは光パルスのピークパワーであり、Ｔ₀はパルス幅（パルスの時間幅）である。このとき、ソリトン次数Ｎは式（３）のように表される。

ソリトンは、ファイバ伝搬と共にＮ＝１を保つようにＴ₀及びＰを変化させる性質を有する。｜β₂｜が長手方向に減少する場合、ソリトンはパルス幅Ｔ₀を｜β₂｜に比例して減少させるとともに、ピークパワーＰをＴ₀に反比例して高める。このようにして、ＤＤＦ中で光パルス圧縮が生ずる。 Here, it is assumed that the optical electric field A (0, t) of the optical pulse at the input end (z = 0) of the DDF is a soliton waveform as shown in the equation (2).

In Equation (2), P is the peak power of the optical pulse, and T ₀ is the pulse width (time width of the pulse). At this time, the soliton order N is expressed as shown in Equation (3).

Soliton has the property of changing T ₀ and P to keep N = 1 with fiber propagation. When | β ₂ | decreases in the longitudinal direction, the soliton decreases the pulse width T ₀ in proportion to | β ₂ | and increases the peak power P in inverse proportion to T ₀ . In this way, optical pulse compression occurs in the DDF.

  By changing the anomalous dispersion value of DDF linearly with respect to the propagation distance z of the soliton, the rate of dispersion reduction acting on the soliton can be kept constant. The details are shown below.

式（４）においてＳは、分散減少の割合を表す定数（スロープ）である。式（４）の微分方程式を解くことにより、式（５）に示すように、ＤＤＦの分散値β₂(z)が得られる。

式（５）は、β₂(0)＜０、Ｓ＞０である場合、図２に示すように、−β₂(z)（又は|β₂(z)|）が入力端から出力端に向けた長手方向に直線的に減少する分散プロファイルであることを示している。このようにして、直線プロファイルにおいて圧縮パルスに作用する分散減少の割合が一定に保たれることがわかる。 Considering that the criterion of the soliton propagation distance z is the dispersion length L _D (z) = T ₀ ² (z) / β ₂ (z), the condition that the rate of dispersion reduction acting on the soliton is constant is It is represented by Formula (4).

In Expression (4), S is a constant (slope) representing the rate of dispersion reduction. By solving the differential equation of equation (4), the dispersion value β ₂ (z) of DDF is obtained as shown in equation (5).

In the formula (5), when β ₂ (0) <0 and S> 0, as shown in FIG. 2, −β ₂ (z) (or | β ₂ (z) |) is changed from the input terminal to the output terminal. It is shown that the dispersion profile decreases linearly in the longitudinal direction toward. In this way, it can be seen that the rate of dispersion reduction acting on the compression pulses in the linear profile is kept constant.

ここで、Ａ(t)及びＡ’(t)は、それぞれ、圧縮前、圧縮後の複素光電界であり、Ｒは光パルスの圧縮率、φは圧縮に伴う位相変化である。 Further, in the DDF having a dispersion profile satisfying the expression (4), a self-similar compression pulse exists as shown below. The self-similar compressed pulse is a pulse that can propagate while maintaining the self-similarity in the DDF having the dispersion profile of Expression (4). By using the self-similar compression pulse as an input pulse of the compression unit 4, optical pulse compression without pedestal generation becomes possible. The relationship of self-similarity is expressed mathematically as Equation (6).

Here, A (t) and A ′ (t) are the complex optical electric fields before and after compression, respectively, R is the compression rate of the optical pulse, and φ is the phase change accompanying the compression.

  The self-similar compressed pulse propagating in the DDF can be obtained by numerical calculation using an averaging method. FIG. 3 shows a flowchart of the averaging method.

First, the initial pulse is set to A (t) (step S1). Next, the pulse waveform A ₁ (t) after compression by DDF is calculated by solving the Schrödinger equation of equation (1) (step S2).

Next, in order to match the phases of A (t) and A ₁ (t), A ₁ (t) exp [-i {argA ₁ (0)}] is newly set to A ₁ (t) ( Step S3). Then, A (t) + R ^−1/2 A ₁ (R ⁻¹ t) is newly set to A (t) (step S4).

Next, in order to keep the energy of A (t) constant, the amplitude of A (t) set in step S4 is adjusted (step S5). Next, returning to step S1, A (t) whose amplitude has been adjusted in step S5 is set as an initial pulse (step S1), and the processes of steps S2 to S5 are repeated. In the processing shown in FIG. 3, A ₁ (t) set in step S 3 is defined as a complex optical electric field A ′ (t) after compression, an optical electric field A (t) before compression, and an optical electric field A after compression. The relationship of '(t) is repeated until the relationship of Expression (6) is satisfied.

FIG. 4 shows the propagation of self-similar compressed pulses in the DDF obtained by the averaging method of FIG. In FIG. 4, in order to generalize the description here, the normalized propagation distance ξ = z / L _D (0) is used as the propagation distance, the nonlinear constant is γ = 1 / W / km, and the dispersion value at z = 0. Is set to β ₂ (0) = − 1 ps ² / km, the slope is set to S = 1/10, and the pulse energy is set to 2 pJ. 4A shows a waveform representing the relationship between the normalized propagation distance, time, and intensity, and FIG. 4B shows a waveform representing the relationship between the normalized propagation distance, frequency, and spectrum. In the case of FIG. 4, the anomalous dispersion value is −β ₂ (ξ) = 1−Sξ = 1−ξ / 10 from Equation (5). Therefore, the anomalous dispersion value and the pulse width T _{0 at} ξ = 5 are half of these values at ξ = 0.

  The optical pulse A ′ (t) compressed by the DDF of the compression unit 4 in this way has a self-similar relationship as shown in the equation (6) with the input pulse A (t). Therefore, the pedestal amount after compression does not increase as compared with the pedestal amount inherent to the self-similar compression pulse before compression. On the other hand, when a pulse compression is performed by propagating an input pulse having a soliton waveform in the DDF of the compression unit 4, a non-soliton component remains along with the compression, so that pedestal is unavoidable.

  FIG. 5 shows the result of calculating self-similar compression pulses obtained with various values of S in order to obtain the optimum S range in Equation (5). 5A shows a waveform representing the relationship between time, intensity (solid line), and instantaneous frequency (dotted line), and FIG. 5B shows a waveform representing the relationship between time and intensity on the logarithmic axis. (C) shows a waveform representing the relationship between frequency and spectrum.

The following knowledge is obtained from the calculation results shown in FIG.
[1] From the instantaneous frequency (dotted line) in FIG. 5A, the self-similar compression pulse is a pulse having an up-chirp.
[2] The smaller S is (that is, the smaller the rate of dispersion reduction), the closer the self-similar compressed pulse is to the soliton waveform.
[3] The self-similar compressed pulse in the case of S <1/8 can be well approximated by a soliton pulse having a linear up-chirp, and the spectrum shape is smooth. These characteristics seem to be stronger as S is smaller, but there is no change that can be discriminated when S <1/12.
[4] In the case of 1/8 <S <1/6, as is clear from FIG. 5B, the intensity waveform has a pedestal. Moreover, it can be seen from FIGS. 5A and 5C that the chirp is not linear and that there is a large distortion in the spectrum. Therefore, when a self-similar compression pulse with 1/8 <S <1/6 is used, pedestal generation is inevitable.
[5] For S> 1/6, self-similar compression pulses are very unstable or cannot exist.

From the above points, it can be seen that high-quality pulse compression with suppressed pedestal generation is possible by using a DDF of S <1/8 and a self-similar compression pulse. On the other hand, if S is made too small, the total dispersion amount of DDF (that is, a value obtained by integrating | β ₂ (z) | in the longitudinal direction) increases. It is known that when pulse compression is performed using such a fiber, pulse timing jitter increases. This is shown in the following literature.
K. Tamura and M. Nakazawa, “Timing jitter of solitons compressed in dispersion-decreasing fibers,” Opt. Lett., Vol. 23, no. 17, pp. 1360-1362, 1998.

For the above reasons, it can be said that the optimum condition in the design of the DDF constituting the compression unit 4 is 1/12 <S <1/8. From the equation (5), considering that the anomalous dispersion value of DDF in the compression unit 4 is halved at a distance 1 / 2S times the dispersion length L _D (0) of the input end (z = 0), this optimum condition Means that the anomalous dispersion value is halved by a distance of 4 to 6 times the dispersion length of the input end.

Note that the numerical calculation shown in FIG. 5 is based on a specific nonlinear constant γ, dispersion value β ₂ (0), and pulse energy, but the above results also apply to other parameters by scaling. Is possible. Also, if it is noted that S is a dimensionless quantity, the above discussion on S is a general discussion that does not depend on scaling.

  Next, the reason why the use of the up chirp pulse is effective in the pulse compression by the DDF of the compression unit 4 will be described with reference to FIGS.

  FIG. 6A shows a time-intensity waveform and frequency-spectrum waveform when a self-similar compression pulse is used, and FIG. 6B shows a time-intensity waveform and frequency-spectrum when a soliton pulse is used. Waveform is shown. As shown in FIG. 6, when a soliton pulse having no chirp (that is, having a constant instantaneous frequency) is used as an input pulse of DDF, a pedestal is generated in a pulse output from DDF. Hereinafter, the cause of the pedestal will be described qualitatively.

  The change in pulse shape in the optical fiber is caused by the dispersion effect. In order for the pulse waveform to be compressed, the input optical pulse has an up chirp, and the up chirp needs to be compensated by anomalous dispersion.

  On the other hand, when an optical pulse without chirp is used as the input pulse, no compression occurs at the input end of the DDF, as shown by the curve a in FIG. After the optical pulse is input, an excessive up chirp due to the Kerr effect occurs as the optical pulse propagates, and the up chirp is compensated by anomalous dispersion, thereby causing pulse compression. That is, a delay in pulse compression occurs with respect to a decrease in the anomalous dispersion value. The excessive Kerr effect at this time causes pedestal generation. Therefore, by applying an up-chirp to the input pulse in advance, pulse compression is generated from the input end of the DDF as shown by the curve b in FIG. 7 (a straight line whose pulse width decreases in proportion to the anomalous dispersion value). Can do.

  If the input pulse has excessive up chirp, excessive pulse compression occurs at the input end of the DDF, as shown by curve c in FIG. That is, the progress of pulse compression occurs with respect to the decrease in the anomalous dispersion value, and in this case also a pedestal occurs.

  As described above, there is an optimum value for the up chirp value of the input pulse. The optimum value of this up-chirp is obtained by a self-similar compression pulse obtained by the averaging method shown in FIG.

ここで、Ａ_０、Ｔ_０、Ｃは、それぞれ、ピーク振幅、時間幅、チャープパラメータを表す。自己相似圧縮パルスは、ＳとＣがほぼ等しい値を取るという性質を有する。従って、Ｓ＜1/8の圧縮部４に入力するパルスとして、Ｓと等しいＣを有するリニアチャープソリトンパルスを使用することもできる。 As shown in [3] above, the self-similar compressed pulse in the case of S <1/8 can be well fitted with a linear chirp soliton waveform as shown in equation (7).

Here, A ₀ , T ₀ , and C represent a peak amplitude, a time width, and a chirp parameter, respectively. The self-similar compression pulse has a property that S and C take substantially equal values. Therefore, a linear chirp soliton pulse having C equal to S can also be used as a pulse input to the compression unit 4 with S <1/8.

<Distributed profile of conversion part>
As in the present invention, when beat light is used as seed light and a short pulse is generated by an optical pulse compression device, beat light cannot be used as input light to the DDF of the compression unit 4. This is because the intensity time waveform of the beat light is a sine wave and is significantly different from the intensity time waveform of the self-similar compressed pulse. For this reason, when beat light is input to the DDF of the compression unit 4, uncompressed components remain and pedestals occur. Therefore, in the optical pulse compression device 2 using the beat light as the seed light, the conversion unit 3 needs to temporarily convert the beat light input from the beat light generation unit 1 into a self-similar compressed pulse.

  The usefulness of the conversion unit that converts beat light into an up-chirp pulse is described in Patent Document 1. However, the conversion unit of the present invention provides a DDF design with higher accuracy than that of Patent Document 1. .

  In the first embodiment, the backward propagation approach is taken to find the dispersion profile of the DDF that constitutes the conversion unit 3 that converts the beat light into self-similar compressed pulses. FIG. 8 shows a schematic diagram of this approach. The optimized DDF of the converter 3 converts the beat light into a self-similar compression pulse train as shown in FIG. 8A. In this approach, as shown in FIG. 8B, self-similar compression is performed. Find the dispersion profile that converts the pulse train into beat light.

式（８）では、図８（ｂ）に示すように、伝播距離ｚ’をＤＤＦの出力端から入力端の方向に定義している。 Hereinafter, a procedure for optimizing the DDF dispersion profile of the conversion unit 3 will be described.
First, the DDF dispersion profile of the conversion unit 3 is expressed by an nth-order polynomial of the propagation distance z ′ as shown in Equation (8).

In Expression (8), as shown in FIG. 8B, the propagation distance z ′ is defined in the direction from the output end of the DDF to the input end.

  Next, the optical pulse A (z ′, t) propagating in the reverse direction (the direction from the output end to the input end) in the DDF having the dispersion profile of Expression (8), and the self-similar compression pulse as the input pulse A ( 0, t) is obtained from the nonlinear Schroedinger equation of equation (1), and the development of an error from the beat light is obtained. The error here refers to time integration of the square of the absolute value of the difference between the complex electric fields of the optical pulse propagating in the opposite direction and the beat light, and normalizing this integrated value with the energy of the optical pulse. Is a value obtained by.

The error between the backward propagating light pulse and the beat light takes a minimum value at a certain propagation distance. Small changing only the m-th order coefficient a _m of the dispersion profile of formula (8), for a local minimum of the error in the backward propagation light pulses and the beat light. By performing this process for each order from 0 to n, by changing the a _m so that the error between the beat light is small, it is possible to obtain an optimum dispersion profile.

FIG. 9 shows the dispersion profile optimized using the backward propagation method and the evolution of errors between the backward propagating light pulse and the beat light. However, the input pulse is a self-similar compressed pulse obtained by the averaging method in the DDF having a linear profile of S = 1/10, and the self-similar compressed pulses are arranged in opposite phases at intervals of 16 ps. The input pulse is used for backward propagation. In this case, as shown in FIG. 9, in the optimal dispersion profile, at the propagation distance z ′ = z ₀ ′ = 11.9 km, the error between the backward propagating light pulse and the beat light is minimal (˜2 × 10 ^{− 5} ) With this propagation distance, 99.998% of the energy of the self-similar compression pulse is converted into beat light. FIG. 10 shows the development of the optical pulse intensity waveform in the optimized dispersion profile. 10 that the self-similar compressed pulse is converted into beat light as it propagates in the opposite direction.

  As described above, in the dispersion profile optimized by using the backward propagation method, the self-similar compressed pulse is converted into beat light. Therefore, by using this optimized dispersion profile in the forward direction, it becomes possible to convert the beat light into a self-similar compressed pulse.

  The dispersion profile optimized using the backward propagation method can be applied to optical pulses of various repetition frequencies and energies by scaling. As an example, FIG. 11 shows a dispersion profile for converting beat light having a repetition frequency of 160 GHz into self-similar compressed pulses. FIG. 11 shows a case where the DDF nonlinear constant γ of the conversion unit 3 is set to 20 / W / km and the intensity of the beat light is set to 250 mW. FIG. 11 also shows a DDF dispersion profile (linear profile) of the compression unit 4 that further compresses the self-similar compression pulse converted by the DDF of the conversion unit 3.

  In the configuration shown in FIG. 11, the variance value and the gradient of the variance value at the output end of the conversion unit 3 are equal to the variance value and the gradient of the variance value at the input end of the compression unit 4. The reason for this can be explained as follows. The conversion unit 3 converts the beat light into solitons and simultaneously compresses the pulses. Therefore, if the dispersion value or the gradient of the dispersion value is discontinuous from the output end of the conversion unit 3 to the input end of the compression unit 4, a compression delay or advance as shown in FIGS. A pedestal occurs. Therefore, in the configuration of FIG. 11, the variance value and the gradient of the variance value at the output end of the conversion unit 3 are equal to the variance value and the gradient of the variance value at the input end of the compression unit 4. By doing in this way, compression in conversion part 3 and compression in compression part 4 occur continuously, and generation of a pedestal can be controlled.

  FIG. 12A shows the intensity waveform of the compression pulse obtained using the dispersion profiles of the conversion unit 3 and the compression unit 4 shown in FIG. As shown in FIG. 12A, a satellite pulse is attached around the main pulse, but its intensity is suppressed by about 50 db as compared with the main pulse, and it can be seen that almost no pedestal is generated. For comparison, FIG. 12B shows the intensity waveform of the compressed pulse obtained only by the compression unit 4 without using the conversion unit 3. Since the beat light input to the compression unit 4 is different from the self-similar compression pulse, a dispersed wave is generated and remains as a pedestal around the compression pulse. As a result, the peak-pedestal ratio can be realized only about 28 db.

  As described above, according to the optical pulse generator 100 of the first embodiment, beat light is converted into a self-similar compressed pulse using the DDF whose anomalous dispersion value decreases in the longitudinal direction. Since the time width is compressed by the DDF whose anomalous dispersion value decreases in proportion to the position in the longitudinal direction of the DDF, the occurrence of pedestal is suppressed, and a compressed pulse with less jitter can be obtained. Can improve the quality.

  In addition, by using beat light as seed light, an ultrahigh frequency signal can be easily created, and optical pulse generation utilizing the ultrahigh frequency signal can be realized.

(Second Embodiment)
Next, a second embodiment of the present invention will be described with reference to FIGS.
In the second embodiment, a CPF (Comb-like Profiled Fiber) in which a nonlinear fiber and a linear fiber are alternately connected is used instead of the DDF constituting the conversion unit 3 and the compression unit 4 of the first embodiment. Yes, the CPF is designed so as to obtain characteristics close to the DDF shown in the first embodiment.

In the second embodiment and the third embodiment described later, the nonlinear fiber and the linear fiber are defined as follows. In the following, the nonlinear constant is γ, and the dispersion value is β ₂ .
Non-linear fiber: γ> 5 / W / km (typical value: γ to 15 / W / km),
−3 <β ₂ <3 ps ² / km (typical value: β ₂ to −1 ps ² / km);
Linear fiber: γ <5 / W / km (typical value: γ-1 / W / km),
β ₂ <-3 ps ² / km (typical value: β ₂ -22 ps ² / km).

  First, the configuration in the second embodiment will be described. FIG. 13 shows a configuration of an optical pulse generator 200 according to the second embodiment. In the optical pulse generator 200 of FIG. 13, the same components as those of the optical pulse generator 100 shown in FIG. 1 in the first embodiment are denoted by the same reference numerals, and the functional description thereof is omitted.

  As shown in FIG. 13, the optical pulse generator 200 includes a beat light generator 1 and an optical pulse compressor 5. The optical pulse compression device 5 includes a conversion unit (converter) 6 and a compression unit (compressor) 7.

  The conversion unit 6 is configured by a CPF in which nonlinear fibers and linear fibers are alternately arranged in the longitudinal direction, converts the beat light input from the beat light generation unit 1 into an up-chirp pulse, and converts the up-chirp pulse into the compression unit 7. Output to. The CPF of the conversion unit 6 is designed to have a dispersion characteristic similar to the DDF of the conversion unit 3 of the optical pulse generator 100 in the first embodiment.

  The compression unit 7 is configured by a CPF in which nonlinear fibers and linear fibers are alternately arranged in the longitudinal direction similarly to the conversion unit 6, compresses the time width of the up-chirp pulse input from the conversion unit 6, and forms an optical pulse train Output. The CPF of the compression unit 7 is designed to have a dispersion characteristic similar to the DDF of the compression unit 4 of the optical pulse generation device 100 in the first embodiment.

Hereinafter, a design method of the CPF constituting the conversion unit 6 and the compression unit 7 will be described.
First, the dispersion value of DDF designed in the optical pulse generator 100 of the first embodiment is β ₂ (z), the nonlinear constant is γ, the length of the DDF in the longitudinal direction is L, and the compression rate of the optical pulse is R And Next, this DDF is divided into N sections, and the length of the mth section is L _m . A method for determining the fiber length L _m of each section will be described later.

First, assuming that the length L _m of each section of the DDF is set, a method for determining the length of the nonlinear fiber and linear fiber CPF corresponding to each section. The CPF in the second embodiment has a configuration in which a plurality of pairs of nonlinear fibers and linear fibers are connected as shown in FIG. Hereinafter, the lengths of the mth nonlinear fiber and linear fiber of the CPF are L _{NL, m} and L _{L, m} , respectively. Further, the nonlinear constant and dispersion value of the mth nonlinear fiber of the CPF are γ _NL and β _{2, NL} respectively, and the nonlinear constant and dispersion value of the mth linear fiber of the CPF are γ _L and β ₂ , respectively. _{, L.}

式（９）において、ｚ_mは、ＤＤＦのｍ番目のセクションの開始位置であり、式（１１）のように定義される。

In the second embodiment, L _{NL, m} and L _L so that the relationship between the mth section of the DDF and the corresponding non-linear fiber and linear fiber of the CPF satisfies the equations (9) and (10). _{, m} is determined.

In Expression (9), z _m is the start position of the mth section of the DDF, and is defined as Expression (11).

Equation (9) represents a condition in which the sum of the dispersion effects in the CPF nonlinear fiber and the linear fiber matches the dispersion effect of the corresponding DDF section. Equation (10) represents a condition in which the sum of the nonlinear effects in the CPF nonlinear fiber and the linear fiber matches the nonlinear effect in the corresponding DDF section. When Equation (9) and Equation (10) are coupled for all sections (m = 1,..., N), 2N simultaneous linear equations are obtained, and L _{NL, m} and L _{L, m} corresponding to each section are expressed as follows. It can be easily calculated.

Next, a method for determining the length L _m of each section of the DDF will be described.
First, when dividing the DDF, it is preferable that the relationship between the pulse width and the section length is the same in each section so that the effect on the optical pulse in each section is equalized. The reason is considered as follows.

When the length of a section with DDF is short, the nonlinear effects and linear effects of the CPF corresponding to the section are small, so that the average of both effects acts on the propagating optical pulse. As a result, CPF and DDF exhibit substantially the same characteristics, and high-quality pulse compression is possible. On the other hand, if only a section having a DDF is lengthened, the nonlinear effect and linear effect of the CPF corresponding to the section increase, and the nonlinear effect and the linear effect act on the optical pulse independently. And DDF show different characteristics. As a result, the quality of the compressed optical pulse is reduced. Therefore, considering that the fiber length, which is the reference for the propagation distance of the optical pulse, is the dispersion length, the length L _m of each section should be shortened with pulse compression.

Furthermore, in the DDF designed according to the first embodiment, the optical pulse self-similarity is maintained throughout the optical pulse compression process, so that each section of the DDF can also be divided in a self-similar manner. Therefore, the length of each section is represented by a geometric sequence using the compression rate R of the optical pulse as shown in Equation (12).

Furthermore, the use of condition that the sum of the length L _m of each section of the DDF is equal to the total length L of the DDF, the length L _m of each section is expressed by the equation (13).

式（１４）において、Ｐ_NL,mはＣＰＦのｍ番目の非線形ファイバを伝播する光の強度、Ｐ_L,mはＣＰＦのｍ番目の線形ファイバを伝播する光の強度であり、Ｐ_mはＤＤＦのｍ番目のセクションを伝播する光の強度である。式（９）及び式（１４）を満たすようにＣＰＦを設計することにより、光損失の影響を排除することができる。 In the above description, the case where the CPF is designed without considering the influence of the optical loss is shown. However, when designing an actual CPF, it is necessary to consider the effect of optical loss. That is, the light intensity in each section of the CPF is different due to the propagation loss in the fiber and the connection loss of different types of fibers. In order to compensate for the effect of optical loss, equation (10) may be changed to the following equation (14).

In Equation (14), P _{NL, m} is the intensity of light propagating through the mth nonlinear fiber of CPF, P _{L, m} is the intensity of light propagating through the mth linear fiber of CPF, and P _m is DDF. Is the intensity of light propagating through the m th section of By designing the CPF so as to satisfy the equations (9) and (14), the influence of optical loss can be eliminated.

  As described above, according to the optical pulse generation device 200 of the second embodiment, the CPF having the same properties as the DDF constituting the optical pulse generation device 100 of the first embodiment is designed. High-quality compressed pulses can be obtained.

(Third embodiment)
Next, a third embodiment of the present invention will be described with reference to FIGS.

  In the second embodiment described above, the CPF characteristics approach the DDF characteristics as the number of pairs of nonlinear fibers and linear fibers that constitute the CPF is increased. On the other hand, as the number of pairs of nonlinear fiber and linear fiber decreases, the nonlinear effect in the nonlinear fiber and the linear effect in the linear fiber inevitably increase. As a result, in the CPF, since the nonlinear effect and the linear effect each independently act on the optical pulse, the characteristics of the CPF are different from the characteristics of the DDF, and the quality of the optical pulse is degraded. That is, there is a trade-off relationship between the number of pairs of nonlinear fiber and linear fiber and the quality of the optical pulse. In the third embodiment, a CPF capable of reducing the amount of pedestal generated while maintaining the number of pairs of linear fibers and linear fibers in the CPF shown in the second embodiment is designed.

  In the third embodiment, the length of the nonlinear fiber arranged at the input end (the start position of the first (m = 1) section) is set to the length of the nonlinear fiber arranged next (m = 2). The CPF is designed to be shorter. In this way, the first section is substantially equivalent to including a non-linear fiber placed in the next section by the length of the non-linear fiber being shortened.

  When the CPF designed in this way is viewed in each section, as shown in FIG. 15 (a), a part of the nonlinear fiber first arranged in the same section (α in the figure). Is placed behind the linear fiber (α ′ in the figure), so that the nonlinear fiber is arranged on both sides of the linear fiber in the same section. In this case, as shown in FIG. 15A, since the dispersion profile in the longitudinal direction has a shape having good symmetry within the same section, it will be referred to as a symmetric CPF hereinafter. In addition, when the CPF in the second embodiment is viewed in each section, as shown in FIG. 15B, a linear fiber is arranged after the nonlinear fiber, and the dispersion profile in the longitudinal direction is within the same section. Since this is an asymmetrical shape, it will be referred to as an asymmetrical CPF hereinafter.

  Next, effects of the symmetric CPF according to the third embodiment will be described. Considering the effect on the optical pulse in a certain section, in the asymmetric CPF, the nonlinear effect always acts on the optical pulse before the linear effect. For this reason, the balance between the nonlinear effect and the linear effect is lost. On the other hand, in the symmetric CPF, the nonlinear effect acts on the optical pulse earlier than the linear effect in a certain section, but the length of the nonlinear fiber acting earlier becomes shorter, so that after the linear effect continues to act. In addition, a non-linear effect acts. For this reason, it is possible to increase the balance between the nonlinear effect acting on the optical pulse and the linear effect in the same section as compared with the asymmetric CPF.

The symmetric CPF is, for example, a nonlinear fiber in which the length L _{NL, 1} of the nonlinear fiber at the input end is arranged second in the CPF with the number of pairs (number of sections) N designed according to the second embodiment. Can be obtained by halving the length L _{NL, 2} .

  FIG. 16 shows a simulation result in the case where the conversion unit 6 uses a CPF with the number of pairs N = 50 and compresses beat light with a repetition frequency of 160 GHz into a pulse train with a pulse width of 100 fs. FIG. 16A shows the intensity waveform of the compressed pulse when the symmetric CPF is used, and FIG. 16B shows the intensity waveform of the compressed pulse when the asymmetric CPF is used. FIG. 16 shows that the use of a symmetric CPF reduces the pedestal around the optical pulse as compared to the case of using an asymmetric CPF.

  FIG. 17 shows the results of examining the peak-pedestal ratio (pedestal / peak) when using a symmetric CPF and using an asymmetric CPF with respect to various values of the number of pairs N. FIG. 17 shows that when the symmetric CPF and the asymmetric CPF having the same number of pairs are compared, the pedestal is reduced by about 6 db when the symmetric CPF is used.

Further, when the pedestal amount is the same, the number N of CPF pairs is smaller when the symmetric CPF is used than when the asymmetric CPF is used. When the reduction rate of the number of pairs in the case of using a symmetric CPF and r, and the asymmetric CPF pair number N, since the pedestal of symmetric CPF pair number rN is equal, from FIG. 17, 22.5-35log ₁₀ (N) = 16.5−35 log ₁₀ (rN) holds. The value of the reduction rate r that satisfies this condition is about 0.67. Therefore, when the symmetric CPF is used, the number of CPF pairs can be reduced to 2/3 than when the asymmetric CPF is used.

  Note that beat light input to the CPF constituting the optical pulse compression device 5 of the second and third embodiments often has a nonlinear effect in the amplification process in the optical amplifier in the beat light generation unit 1. And undergoes a linear effect. If the CPF is designed without taking these effects into consideration, the compression process in the CPF is affected, and the pedestal and the like may increase. In order to eliminate this influence, the CPF may be designed by regarding the optical amplifier in the beat light generator 1 as the first (input end) nonlinear fiber of the CPF.

  As described above, according to the third embodiment, the length of the nonlinear fiber at the input end (m = 1) of the CPF constituting the optical pulse compression device 5 of the second embodiment is set to the next (m = By designing it to be shorter than the length of the nonlinear fiber arranged in 2) (for example, half the length), generation of pedestals can be further reduced, and a higher quality compressed pulse can be obtained.

(Fourth embodiment)
Next, a fourth embodiment of the present invention will be described with reference to FIGS.
In the fourth embodiment, an optical pulse generation experiment performed using the CPF designed based on the first to third embodiments will be described.

  FIG. 18 shows configurations of the beat light generator 10 and the converter 11 in the pulse generator of the fourth embodiment. The beat light generator 10 includes two DFB-LDs (distributed feedback laser diodes) 12 and 13, a polarization maintaining coupler 14, and an EDFA (erbium-doped fiber amplifier) 15.

  In the beat light generation unit 10, the continuous light output from the DFB-LDs 12 and 13 is combined by the polarization maintaining coupler 14, and the combined light is amplified by the EDFA 15 to generate beat light. The repetition frequency, center wavelength, and average power of the beat light are 160 GHz, 1555 nm, and 500 mW, respectively.

  The conversion unit 11 is configured by a CPF in which a highly nonlinear fiber (HNLF) and a single mode fiber (SMF) are alternately arranged, and converts the beat light input from the beat light generation unit 10 into an up-chirp pulse. The CPF applied in the fourth embodiment is composed of 40 pairs of HNLF and SMF.

FIG. 19 shows a CPF dispersion profile (upper stage) and a non-linear profile (lower stage) in the fourth embodiment. The nonlinear constant γ of the CPF HNLF is 14.2 / W / km, and the average value of the dispersion β ₂ is 0.32 ps ² / km. In designing the CPF in the fourth embodiment, the dispersion profile (see FIG. 11) obtained in the first embodiment is simulated by the CPF based on the guidelines of the second and third embodiments. The length of the fiber is determined. The total length and optical loss of this CPF are 684 m and 4.1 dB, respectively.

FIG. 20 shows the measurement result of the CPF output pulse in the fourth embodiment. As shown in FIG. 20A, the intensity autocorrelation waveform matches a fitting curve assuming a seq ² pulse with a full width at half maximum of 324 fs. The measured value of the peak pedestal ratio is limited by the measurement system and is 21 dB. From FIG. 20 (b), it can be seen that the spectrum closely matches the sech ² curve with a bandwidth of 1.0 THz, and no distortion of the spectrum occurs. Note that the time bandwidth product is 0.32.

  According to the fourth embodiment described above, an extremely high purity optical soliton array can be obtained by using the CPF designed based on the guidelines of the first to third embodiments.

(Fifth embodiment)
Next, a fifth embodiment of the present invention will be described with reference to FIGS.
In the fifth embodiment, a terahertz repetitive optical pulse generation experiment performed using the CPF designed based on the first to third embodiments will be described.

  FIG. 21 shows the configuration of the beat light generator 20 and the converter 21 in the pulse generator of the fifth embodiment. The beat light generation unit 20 includes two DFB-LDs (distributed feedback laser diodes) 22 and 23, two OBPFs (Optical Band Pass Filters) 24 and 25, a polarization maintaining coupler 26, and an EDFA (erbium). Additive fiber amplifier) 27.

  In the beat light generation unit 20, light having a specific wavelength is extracted by the OBPFs 24 and 25 from the continuous light respectively output by the DFB-LDs 22 and 23, and the two extracted lights are combined by the polarization maintaining coupler 26. Then, the synthesized light is amplified by the EDFA 27 to generate beat light. The repetition frequency, center wavelength, and average power of the beat light are 1.0 THz, 1550 nm, and 1.25 W, respectively.

The conversion unit 21 is composed of a CPF in which HNLF and SMF are alternately arranged, and converts the beat light input from the beat light generation unit 20 into an up-chirp pulse. FIG. 22 shows the relationship between the number of fibers of HNLF and SMF and the fiber length. In FIG. 22, L _HNLF indicates the fiber length of HNLF, and L _SMF indicates the fiber length of SMF. The CPF applied in the fifth embodiment is composed of 15 pairs of HNLF and SMF. The nonlinear constant of HNLF is 14 / W / km, and the average value of the dispersion is 0.54 ps ² / km. The method for determining the length of each fiber is the same as in the fourth embodiment. The total length and optical loss of this CPF are 115 m and 1.9 dB, respectively.

23 and 24 show the measurement results of the CPF output pulse in the fifth embodiment. FIG. 23A shows an intensity autocorrelation waveform in a scan range of 5 ps. FIGS. 23B and 23C are an enlarged view of a central portion of one waveform of the intensity autocorrelation waveform and a logarithmic plot thereof, respectively. The figure is shown. The intensity autocorrelation waveform shown in FIG. 23 agrees well with the fitting curve assuming a seq ² pulse with a full width at half maximum of 97 fs. The upper part of FIG. 24 shows the spectrum of beat light having a repetition frequency of 1.0 THz, and the lower part of FIG. 24 shows the spectrum of the output pulse of CPF. As shown in the lower part of FIG. 24, the spectrum of the CPF output pulse spreads over 100 nm and its full width at half maximum is 3.5 THz. The time bandwidth product is 0.34.

  According to the fifth embodiment described above, the CPF design guidelines shown in the first to third embodiments can be applied to the generation of a terahertz repetitive ultrashort optical soliton array, and the optical soliton having an extremely high purity. A row can be obtained.

1 is a block diagram showing a configuration of an optical pulse generator according to a first embodiment of the present invention. The figure which shows the dispersion | distribution profile of the compression part (DDF) of FIG. The flowchart which shows the process which calculates a self-similar compression pulse using the averaging method. The figure which shows the numerical calculation result of propagation in DDF of a self-similar compression pulse. The figure which shows the self-similar compression pulse for various S (slope in the longitudinal direction of dispersion). The figure which shows the time-intensity waveform and frequency-spectrum waveform (a) of a self-similar compression pulse, and the time-intensity waveform and frequency-spectrum waveform (b) of a soliton pulse. The figure which shows the relationship between the propagation distance of an optical pulse, an anomalous dispersion value, and a pulse width. The figure which shows the forward propagation (a) and the backward propagation (b) in the conversion part (DDF) of FIG. The figure which shows the evolution of the difference | error of the dispersion | distribution profile optimized by the backward propagation method, and a self-similar compression pulse and beat light. The figure which shows the mode of reverse propagation (self-similar compression pulse-> beat light). The figure which shows the dispersion | distribution profile of the conversion part of 1st Embodiment, and a compression part. The pulse intensity (a) obtained using the conversion unit of the first embodiment and the pulse intensity (b) obtained using a DDF having a dispersion profile in which the dispersion value decreases linearly with distance are shown. Figure. The block diagram which shows the structure of the optical pulse generator which concerns on the 2nd Embodiment of this invention. The schematic diagram of DDF of 1st Embodiment, and CPF of 2nd Embodiment. The figure which shows the anomalous dispersion value (a) of symmetric type CPF, and the anomalous dispersion value (b) of asymmetric type CPF. The figure which shows the compression pulse (a) at the time of using symmetric type CPF, and the compression pulse (b) at the time of using asymmetric type CPF. The figure which shows the pair number dependence of the peak pedestal ratio of the pulse compressed by CPF. The figure which shows the structure of the beat light generation part and conversion part in the pulse generator of 4th Embodiment. The figure which shows the dispersion | distribution profile (upper stage) and nonlinear profile (lower stage) of CPF in 4th Embodiment. The figure which shows the measurement result of the output pulse of CPF in 4th Embodiment. The figure which shows the structure of the beat light generation part and conversion part in the pulse generator of 5th Embodiment. The figure which shows the relationship between the number of fibers of a highly nonlinear fiber (HNLF) and single mode fiber (SMF), and fiber length in 5th Embodiment. The figure which shows the measurement result of the output pulse of CPF in 5th Embodiment. The figure which shows the measurement result of the output pulse of CPF in 5th Embodiment.

Explanation of symbols

1, 10, 20 Beat light generation unit 2, 5 Optical pulse compression device 3 Conversion unit (DDF)
4 Compression unit (DDF)
6,11,21 Conversion unit (CPF)
7 Compression unit (CPF)
100,200 optical pulse generator

Claims (11)

In an optical pulse compression device that inputs beat light and compresses the time width of the beat light to generate an optical pulse,
An anomalous dispersion value decreases in the direction from the input end to the output end, and a converter that converts the beat light into an up-chirp pulse,
An anomalous dispersion value decreases in proportion to the position in the direction from the input end to the output end, and a compression unit that compresses the time width of the up-chirp pulse,
An optical pulse compression device comprising:   2. The optical pulse compression device according to claim 1, wherein the anomalous dispersion value is halved at a distance of 4 to 6 times the dispersion length at the input end in the compression unit.   The optical pulse compression device according to claim 1, wherein the conversion unit and / or the compression unit is a dispersion reducing fiber.   3. The optical pulse compression device according to claim 1, wherein the conversion unit and / or the compression unit is a comb-shaped profile fiber in which nonlinear fibers and linear fibers are alternately connected.   5. The optical pulse compression device according to claim 4, wherein the comb-shaped profile fiber has a length in a longitudinal direction of a nonlinear fiber arranged at an input end shorter than a nonlinear fiber arranged next. 6.   An optical pulse generator comprising: a beat light generator that generates beat light; and the optical pulse compressor according to claim 1. Generates beat light,
The beat light is converted into an up chirp pulse using a converter in which the anomalous dispersion value decreases from the input end to the output end,
An optical pulse compression method, wherein a time width of the up-chirp pulse is compressed using a compressor whose anomalous dispersion value decreases in proportion to a position in a direction from an input end to an output end.   8. The optical pulse compression according to claim 7, wherein the time width of the up-chirp pulse is compressed using a compressor whose anomalous dispersion value is halved at a distance of 4 to 6 times the dispersion length at the input end. Method.   9. The optical pulse compression method according to claim 7, wherein the converter and / or the compressor is a dispersion reducing fiber.   9. The optical pulse compression method according to claim 7, wherein the converter and / or the compressor is a comb-shaped profile fiber in which nonlinear fibers and linear fibers are alternately connected.   11. The optical pulse compression method according to claim 10, wherein the comb-shaped profile fiber has a length in a longitudinal direction of a nonlinear fiber arranged at an input end shorter than a nonlinear fiber arranged next.

Images