JPH10339687A - Method for evaluating multimode optical fiber - Google Patents

Abstract

PROBLEM TO BE SOLVED: To evaluate a multimode optical fiber including the effect of mode dispersion by fitting Gaussian function to the waveform of an optical pulse through superposition of Gaussian functions of same number as the propagation modes of the multimode optical fiber. SOLUTION: An optical pulse from a common light source 1 is divided through a photocoupler 9a into two which pass through an FC connector 8a and an MPO connector 7a and enter into a specified multimode optical fiber 2 of a four core optical fiber tape 4. The propagated optical pulse is observed by a light receiving unit 3 through a photocoupler 9b. In this regard, the waveform of optical pulse from the light source 1 operating with a pulse signal from a pulse generator 6 can be reproduced by superposing five Gaussian functions when the number of propagation modes is five, for example. Central position of each Gaussian function corresponds to the time lag of each mode light and the magnitude of each Gaussian function represents the intensity of each mode light. Consequently, distribution of mode light can be measured for the optical fiber 2 from the superposed Gaussian functions. According to the method, the optical fiber can be evaluated including the effect of mode dispersion.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The method for evaluating a multimode optical fiber according to the present invention relates to an optical fiber array used for signal transmission between a communication device and a computer. It can be used to objectively evaluate reliability and the like, and is particularly suitable for evaluating the delay time difference (skew) of propagating light.

[0002]

2. Description of the Related Art In recent years, optical parallel transmission has attracted attention from the viewpoint of large-capacity transmission and high-speed processing of information. Among them, an optical interconnect using an optical fiber is expected to improve the performance of a wiring technique between computer systems.
According to this technology, high-speed and high-density wiring can be realized with a light weight and a small diameter by replacing a conventionally used bundle of coaxial cables with an optical fiber array (consisting of a plurality of optical fibers). it can. The basic method of the optical interconnect is a multi-channel parallel synchronization using an electric / optical conversion array (E / O) and an optical / electric conversion array (O / E) and an optical fiber array (Fiber array) as shown in FIG. This is a transmission method, and it is important that electric signals input at the same time in each channel be output almost simultaneously on the output side of the electric signals. Considering a system using an optical interconnect as a whole, it is generally assumed that a delay time difference (skew) between signal channels allowed in the system is about 1 ps / m.

[0003] The definition of the skew of an optical fiber is as follows.
While the cause of the delay is strictly defined in a single mode optical fiber where only chromatic dispersion is involved, in a multimode optical fiber the phenomena becomes complicated due to mode dispersion and mode coupling, so a strict definition is still required. Not fixed
At present, there is no objective evaluation method. However, in the industrial world, the system using multi-mode optical fiber has been studied and put into practical use, mainly in the United States, because of its low cost and easy electrical / optical and optical / electrical coupling. Development has begun, and it is unknown at this time how the definition of skew in multimode optical fiber is ultimately determined by a public agency.However, as the system is under development, Due to the compatibility and reliability of the optical fiber, a means for objectively evaluating the skew of the multimode optical fiber is urgently required.

[0004] There are two typical skew evaluation / measurement methods that are currently performed: 1) a phase method, and 2) a pulse method (also called a propagation method).

[0005] 1) The phase method is to propagate a sine-wave modulated light to an optical fiber to be measured and measure the phase of the propagated light. Two optical fibers (one is a reference fiber, Sine wave modulated light oscillated from a common light source is propagated at the same time, and the waveforms of the sine wave modulated light propagating through these optical fibers are examined on the same time axis, and their phase angles are measured. This is a method of measuring the delay time difference between optical fibers by measuring the difference. FIG. 5 shows an outline of a measuring apparatus for this phase method. A sine wave electric signal generated by a signal generator A is applied to a laser (light source) B to generate a sine wave modulated light. Then, the light is propagated to two optical fibers D to be measured, and the sine wave waveform of each propagated light is observed by the oscilloscope F via the O / E probe E. Then, by examining the phase difference between the two pulse waveforms observed by the oscilloscope F, the phase angle difference between the optical fibers D is obtained, and the skew is derived therefrom.

2) In the pulse method, a light pulse is propagated through an optical fiber to be measured, and the pulse and another reference pulse (trigger electric signal) are checked on the same time axis. This is a method of measuring a delay time difference. FIG. 6 shows an outline of a measuring apparatus for the pulse method, in which a signal generated by a pulse generator A is output to an E / O converter (light source) B and an optical oscilloscope (sampling oscilloscope) G, and the E / O is output. The optical pulse generated by the converter B and propagated through the optical fiber D to be measured and the trigger signal from the pulse generator A are compared and observed by an optical oscilloscope (sampling oscilloscope) G, so that the two can be measured on the time axis. Then, the delay time of the light propagating through the optical fiber D is measured from the position information at step (1). Further, replace the optical fiber D with another one and measure the delay time for the trigger signal in the same manner,
Optical fiber D measured first and optical fiber D measured later
Is derived with reference to a trigger signal which is a common reference.

In the pulse method, in order to derive a delay time difference from an observed waveform, a standard for objectively determining the position on the time axis of the observed waveform is required. Up to now, since the incident pulse waveform has a Gaussian type distribution (normal distribution) (an example of an actual pulse waveform is shown in FIG. 7), the output waveform has the same single Gaussian number (y = Aexp (-((xx ₀ ) / a) ² ) was applied, and the center position of the Gaussian function was used as the position of the waveform, and the delay time in the optical fiber A was Δτ _A , and the delay time in the optical fiber B was Δτ _{When B} , the difference Δτ ′ Δτ ′ = | △ τ _A − △ τ _B | (1) was taken as the skew between the optical fibers.

[0008]

The method of estimating skew by applying a single Gaussian function to a waveform in the pulse method described above was effective for a single mode optical fiber having no mode dispersion. However, there is a problem that the multimode optical fiber is not effective for the reasons described below.

Since a multi-mode optical fiber has a plurality of propagation modes, an optical pulse incident from a light source is
Since the power is distributed to multiple modes and propagates through the optical fiber at different propagation speeds for each mode, the waveform of the output optical pulse does not collapse into a single mountain shape as in a single mode optical fiber. It will be.
In addition, the output waveform changes the distribution of optical power depending on the operating state of the light source and the connection between the light source and the optical fiber, and causes mode coupling while the excited mode light propagates through the optical fiber. Thus, it becomes more ambiguous and deformed, making it difficult to apply a single Gaussian function to the output waveform. Even if the output waveform has a single mountain shape, it often cannot always be said that it measures an accurate skew.

For example, among the modes excited by the optical fiber A, the power of the incident pulse is distributed most to the lowest-order mode (hereinafter referred to as mode 1), and the influence of the distribution to other modes can be ignored. Assuming that the same situation occurs in the optical fiber B, the skew Δτ ′ between the optical fibers due to the fitting of the Gauss function is Δτ ′ = the difference between the arrival times in the mode 1 between the fiber A and the fiber B.・ (2) Further, although the fiber A is in the above-mentioned situation, when the excitation in another mode 2 becomes dominant in the fiber B, the skew Δτ ′ between the optical fibers due to the fitting of the Gauss function becomes Δτ ′. = The difference between the arrival time of fiber A in mode 1 and the arrival time of fiber B in mode 2 (3)

In any of the above patterns, the apparent output waveform is distributed in a single mountain shape. In the above equation (3), the skew between mode 1 and mode 2 is evaluated, and (2) In the expression, the skew between the modes 1 is evaluated.

As described above, in the multimode optical fiber, the output waveform is deformed due to the mode distribution of the optical power, the operation state of the light source, the connection state between the light source and the optical fiber, and the like, so that a single Gaussian function is obtained. Cannot be applied, or even if the output waveform looks like a single Gaussian distribution,
The problem is that accurate skew of the optical fiber cannot be measured because the skew between different modes is evaluated. It is also important to note that this problem involves factors other than the effect of the optical fiber itself, even if the measurement is performed using the same optical fiber and the measurement conditions performed by the operator differ slightly. That is, the skew value is easily evaluated differently.

[0013]

SUMMARY OF THE INVENTION In the present invention, a precise definition of the skew of a multimode optical fiber has not yet been provided at the present stage, and a precise definition of the skew will be provided in the near future. If so, keep in mind that in such a case, the above-mentioned problems must be given in a resolved form, and how much light is distributed between the modes, It is an object of the present invention to provide a method capable of evaluating how much a delay time difference occurs between them.

In the method for evaluating a multimode optical fiber according to the first aspect of the present invention, an optical pulse is input to a multimode optical fiber 2 to be measured as shown in FIGS. A method for evaluating a multi-mode optical fiber by applying a Gaussian function to a waveform of an optical pulse propagated through the optical fiber 2, comprising:
By applying the Gauss function to the same number of Gauss functions as the number of propagation modes of the multi-mode optical fiber 2, the multi-mode optical fiber is evaluated in a form including the influence of mode dispersion. Things.

In the method for evaluating a multimode optical fiber according to the second aspect of the present invention, Ga applied to an optical pulse waveform may be used.
The center position of the uss function is shifted corresponding to the delay time for each mode of the multimode optical fiber 2.

According to a third aspect of the present invention, there is provided a method for evaluating a multimode optical fiber according to the first or second aspect, wherein the plurality of Gaussian functions applied to the optical pulse waveform are used. The distribution of the optical power is evaluated by evaluating the magnitude of the optical power.

According to a fourth aspect of the present invention, there is provided a method for evaluating a multimode optical fiber, wherein the evaluation method according to any one of the first to third aspects is performed for a plurality of multimode optical fibers to be measured. The skew and the distribution of optical power between the multi-mode optical fibers 2 are evaluated by comparing the results in the same mode.

According to a fifth aspect of the present invention, there is provided a method for evaluating a multimode optical fiber, wherein the plurality of multimode optical fibers to be measured are subjected to the evaluation method according to the first to third aspects. If no significant difference is found even when the results are compared between the same modes, the skew and the distribution of optical power between the multimode optical fibers 2 are evaluated by focusing only on the desired one mode. Is what you do.

[0019]

FIG. 1 shows a delay difference measuring system for a multi-core GI optical fiber array for a large-scale computer optical interconnect, which is configured to carry out the method for evaluating a multimode optical fiber according to the present invention. This is an example in which a skew measurement is performed by using an 850 nm laser diode LD as a light source (common light source) 1 and transmitting an optical pulse to each optical fiber 2 of a four-core optical fiber tape 4. In this measurement system, the light pulse of the light source 1 is split into two by an optical coupler 9a, and each of the light pulses enters a desired optical fiber (multi-mode optical fiber) 2 of a four-core optical fiber tape 4 through an FC connector 8a and an MPO connector 7a. The optical pulse propagated through the two optical fibers 2 is an MPO.
The light can be observed by the light receiver (sampling oscilloscope) 3 through the connector 7b, the FC connector 8b, and the optical coupler 9b.

The light source 1 which operates by receiving a pulse signal from the pulse generator 6 shown in FIG. 1 is a short light pulse as shown in FIG. 2, and has a substantially Gaussian distribution shape having a full width at half maximum of about 160 ps. I have. Points in FIG. 2 indicate actually measured data. This optical pulse waveform can be reproduced by superposition of five Gaussian functions whose centers coincide with each other (because the number of propagation modes of the optical fiber 2 is five) as shown in the figure. The solid line is the waveform when the five Gaussian functions are superimposed, which is in agreement with the measured value indicated by the dot.

Next, the light propagated through the optical fiber 2 (10
FIG. 3 shows the waveform of the output pulse observed by the sampling oscilloscope 3 after 0 m). In this case, the output pulse waveform can be reproduced by superimposition of five Gauss functions shifted from the center. Here, the center position of each Gauss function corresponds to the delay time of each mode light, and
The magnitude of the Gauss function indicates the intensity of each mode light.
Therefore, the mode distribution of the mode light of the optical fiber 2 can be checked from the superposed Gaussian function, and the same measurement is performed for the other remaining optical fibers 2 in the four-core optical fiber tape 4 to obtain the mode. The skew and mode light distribution can be evaluated in a form including the influence of dispersion.

The refractive index distribution of the individual optical fibers (multimode optical fibers) 2 constituting the four-core optical fiber tape 4 is of a parabolic type (α = 2). In this case, it is known that the number of propagation modes of the signal light having the wavelength λ is the largest integer less than or equal to N given by the following Equation 1, and the total number of modes of the fiber 2 can be specified.

[0023]

(Equation 1)

Here, a is the core radius, and λ is the wavelength of the signal light. n ₁ is the maximum refractive index of the cladding, and △ is the relative refractive index difference.

By substituting the physical property values a = 25 μm, n ₁ = 1.467, and △ = 1.86% of the actual parabolic optical fiber 2 into the above-mentioned variables and using the signal light wavelength of 850 nm, the total number of modes becomes N = 5.
It becomes 042. Therefore, the number of existing propagation modes becomes five.

The delay time of each mode when the optical fiber 2 propagates 100 m can also be obtained by calculation, and the waveform fi is calculated from the number of modes and the delay time for each mode.
Five Gaussian functions to use can be created.

Next, it was verified whether the evaluation method of the present invention can perform an appropriate evaluation. Fitting was performed on the two optical fibers 2 using the multiple Gauss functions described above, and the group velocities of the highest-order mode and the lowest-order mode in each optical fiber 2 were examined. In this case, the skew compared in the same mode was 1.34200 ps / m when calculated between the fast modes, and 1.34218 ps / m when the slow modes were selected. The result of skew measurement performed by selecting the fast mode and the slow mode was 2.16400 ps / m. From this, it was found that the difference in the skew obtained as a result of the measurement between the same modes was 0.00018 ps / m, which was sufficiently negligible. In contrast, the difference in skew between different modes is 0.822p
s / m. Therefore, this value of 0.822 ps / m is a skew caused by the comparison between different modes, and is considered to be an effect of mode separation. That is, in the evaluation method according to the present invention, one
Information that could not be obtained by Gaussian function approximation can be obtained. That is, the effect of the mode dispersion and the effect of the delay time difference in the same mode can be measured separately.

In order to analyze an optical pulse that is mode-dispersed by propagating through the multimode optical fiber 2 by applying the Gaussian function according to the method of the present invention, the number of Gaussian functions to be superimposed and the position of each Gaussian function (delay time difference) ), Which can be determined from the refractive index distribution of the optical fiber 2. Hereinafter, a general example based on the WKB method analysis will be briefly described.

The refraction profile n (r) of the multimode optical fiber is considered as an α-th power distribution fiber represented by the following equation.

[0030]

(Equation 2)

A: core diameter n ₁ : refractive index n at the center of the core
₂ : Refractive index of cladding

In this case, the number of propagation modes and the delay time are WKB
It is known that it is obtained by calculation by the method. For example, the total number N of propagation modes in the α-th power distribution fiber can be expressed as in Equation 3.

[0033]

(Equation 3)

The group delay time τν of the ν-th mode of the α-th power distribution fiber can be matched by the following equation (4).

[0035]

(Equation 4)

Is given by Here, χ is calculated using the mode number ν, the total number of propagation modes N, and the refractive index distribution index α.

[0037]

(Equation 5)

Is 0 in the lowest order mode and 1 in the highest order mode. Also, y is

[0039]

(Equation 6)

Is a correction term of α called “profile variance”.

[0041]

According to the method for evaluating a multimode optical fiber of the present invention, the following effects can be obtained. 1. When performing a skew measurement of a multimode optical fiber by the pulse method, by approximating the measured waveform with a plurality of Gaussian waveforms corresponding to the total number of modes of the optical fiber 2, the conventional approximation of a single Gaussian waveform can be performed. could not,
An evaluation excluding the influence of mode dispersion can be performed. 2. It is possible to evaluate the delay time due to the same mode between different optical fibers 2. 3. Factors causing skew can be separated and analyzed precisely.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing a measurement system for a method for evaluating a multimode optical fiber according to the present invention.

FIG. 2 shows Gaussian light pulses of a light source in the measurement system of FIG.
FIG. 4 is an explanatory diagram showing a state in which functions are superimposed.

FIG. 3 is an explanatory diagram showing a state in which a Gaussian function is superimposed on an output light pulse in the measurement system of FIG. 1;

FIG. 4 is an explanatory diagram of an optical interconnect.

FIG. 5 is an explanatory diagram showing a state of skew measurement by a conventional phase method.

FIG. 6 is an explanatory diagram showing a state of skew measurement by a conventional pulse method.

FIG. 7 is a waveform diagram of light pulses of a light source in the pulse method of FIG.

[Explanation of Signs] 2 Optical fiber

 ──────────────────────────────────────────────────の Continued on the front page (72) Inventor Haruki Ogoshi 2-6-1 Marunouchi, Chiyoda-ku, Tokyo Furukawa Electric Co., Ltd.

Claims (5)

[Claims] 1. A multi-mode optical fiber to be measured.
And applying a Gaussian function to the waveform of the optical pulse propagated through the multimode optical fiber (2) to evaluate the multimode optical fiber. Apply the Gauss function
As many as the number of propagation modes of the multimode optical fiber (2)
A method for evaluating a multi-mode optical fiber, comprising: evaluating a multi-mode optical fiber in a form including the influence of mode dispersion by superimposing Gaussian functions. 2. A multi-mode optical fiber according to claim 1, wherein the center position of the Gaussian function applied to the optical pulse waveform is shifted in accordance with the delay time of each mode of the multi-mode optical fiber. Evaluation method of optical fiber. 3. The method for evaluating a multimode optical fiber according to claim 1, wherein the distribution of optical power is evaluated by evaluating the magnitudes of a plurality of Gaussian functions applied to the optical pulse waveform. A method for evaluating a multimode optical fiber, comprising: 4. The multimode optical fiber (2) to be measured is subjected to the evaluation method according to any one of claims 1 to 3, and the results are compared between the same modes to obtain a multimode optical fiber. A method for evaluating a multi-mode optical fiber, comprising: evaluating a skew between optical fibers (2) and an optical power distribution. 5. The evaluation method according to claim 1 is performed on a plurality of multimode optical fibers to be measured, and the results are compared in the same mode. A method for evaluating a multi-mode optical fiber, comprising: evaluating a skew and an optical power distribution between the multi-mode optical fibers (2) by focusing on only one desired mode when no difference is found.