Optical pulse reflectometry device and method

The present invention relates to optical fiber sensors based on optical time domain reflectometers, namely the measurement of the reflection coefficient (Oikawa) at which the pulse or series of pulses are introduced into the fiber and the signal returned to the input end and is generated by the reflected and scattered light in the fiber.

One example of such sensors is the Raman Oikawa sensor with which can be determined by the temperature profiles of the intensity distribution of the return signal as a function of time measured from the moment the test pulse and, similarly, the distance along the fiber. Oikawa technology is extremely widely used in telecommunications to assess the quality and integrity of the fiber-optic communication lines. In this case, the line connection is determined by the requirements of the communication system and Oikawa must be designed to interface with a fiber optic system. However, in applications for sensors Oikawa there is much more flexibility when choosing a fiber-optic system and its interconnections to meet the requirements of the developer Oikawa sensors.

It is known that the performance of sensors Oikawa limited power which can be fed into the fiber due to the presence of non-linear optical effects. Source of said non-linear optical effects varies depending upon the characteristics of the test pulse. For short pulses having a large spectral width, the predominant effect is usually stimulated Raman scattering (CDF). For longer pulses with a spectral width of less prevalent effect is the total stimulated Brillouin scattering (FBG). For narrow-band pulse limiting effect it is generally a self phase modulation (SPM) in cases where the momentum required to support small spectral width. A more detailed explanation of these phenomena can be found in the literature, for example, GP Agrawal "Nonlinear fiber optics", Academic Press 1995 ISBN 0-12-045142-5.

The limited power that can be fed into the optical fiber for measurement type Oikawa, is a serious restriction of the performance of these systems. The signals received by the test pulse in these systems are usually very weak; typically only a few photons are returned in the return signal from the most remote points of interest in the system. Since the received signal is proportional to the power of the probe pulse it is obvious that the possibility of increasing the power of the probe pulse leads to higher productivity. Alternatively, if the energy is returned for this introduced power can be increased, then it may be increased signal-to-noise ratio and thus improve the quality of measurement. Finally, it should be noted that the measurements Oikawa usually averaged over a large number of measurements to improve the signal to noise ratio, the latter value is improved as the square root of the averaged measurements. This improvement in signal to noise ratio obtained at a higher power supplied may be used to obtain a shorter time updates the measurement.

Therefore, necessary to provide methods and devices for improving the signal-to-noise measurement Oikawa fiber segment of interest when such a segment is separated from the device finite distance. It often happens that the most important segment is at the outermost end of the fiber. While it may require a low fiber length measurement, the present invention is directed to improving the quality of remote measuring fiber length.

The main reason for concentrating on improving measurement at the remote end of the fiber is that in this area the attenuation undergone by a probe pulse in the outbound direction and the signals in the reverse direction is the highest. However, there are applications in which the end-section is major interest and in which the quality of data at short distances is less important. One example of such an application is the measurement of temperature profiles in offshore oil wells. In this case it is important to know the temperature of the fluid flowing in the well, typically up to 0.1 ° C. However, the equipment is usually located on the platform, which is placed at a certain distance from the well, wherein the well is connected to the platform by means of an underwater pipeline that lies (or recessed) on the seabed, the riser delivers oil from an underwater pipeline up to the platform. The horizontal distance between the platform and the mouth of an underwater well known as a distance "removal." In a typical example of wells intended for deep water oil production, the well might extend up to ten kilometers below the seabed. Separation distance can be up to 20-30 km and can reach a depth of 2000 m and in the future may exceed this value. In this example it is the segment of optical fiber downhole end 10 km is important. The fiber that connects the platform to the wellhead, is less important: it can give information about blockage of the underwater pipeline, but the temperature resolution required for the purpose of guaranteeing the flow in subsea pipelines is much less demanding than that required in the well.

For additional disclosure discusses the concept of the present invention, the relationship between the maximum power that can be fed into the fiber and the resulting backscatter power.

The resolution of the measured value (eg temperature) Oikawa type sensors, such as Raman and Brillouin Oikawa Oikawa, is generally determined by the ratio of signal to noise ratio of the signal backscatter. For further discussion of this can be found in the handbook "Optical Fibre Sensor Technology" edited by Grattan and Meggit (Chapman & Hall, London, 1995, ISBN 0 412 59210 X), and especially in the chapter on the distributed fiber optic sensors. Summing up, the return signal in such a sensor is proportional to the energy of the probe pulse. To increase the pulse energy may be increased or the pulse duration or the pulse power. In the first case, the spatial resolution (ie, the ability of the sensor to distinguish between closely spaced features of the profile of the measured value) is deteriorating. In the second case, the peak power is limited by non-linear effects which convert the probe pulse at a wavelength different from the input, if the power exceeds certain limits.

In general, it is the intensity of the optical power in the fiber determines the occurrence of nonlinear effects. Intensity is the ratio of the optical power supplied to the area through which said power is distributed. Since the optical power is not homogeneous, "effective" area is generally defined as follows:

<img he="24" wi="39" file="00000002.tif" img-content="undefined" img-format="tif" />

where Ψ (r) is the electric field distribution as a function of the radial coordinate r. The scale of the nonlinear effects is inversely proportional to Aeff (effective area). From this it follows that for a given limit at which non-linear effects become unacceptable, the power which can be fed into the fiber, Aeff increases proportionally. However, it was found that the backscatter factor, i.e. power ratio of the backscatter signal to the energy of the probe pulse, is inversely proportional to Aeff. It follows that, if the configuration of the fiber varies to increase the Aeff, a fiber can be fed more power, but part of the pulse energy which is converted into the feedback signal is reduced approximately in proportion to increase Aeff, which leads to the fact that the return signal is unchanged. While dopants used to modify the refractive index leads to the fact that the said dependence in some way different from proportionality with Aeff, the foregoing discussion remains valid, at least approximately. While some effects are scaled further. For example, for small additions of GeO2 to silica, the refractive index of the core increases in proportion to the molar concentration of GeO2, which leads to the fact that the effective area decreases in inverse proportion to the square of the concentration of GeO2. However, the threshold for stimulated Raman scattering is reduced in the same proportion as the Raman amplification and the cross section and the spontaneous Raman scattering (which results in a stimulated Raman scattering) are both proportional to GeO2 concentration.

In the previous paragraph uses terms that relate to single mode fibers; but it is clear that the same principles apply to multimode fibers.

According to a first aspect of the present invention there is provided an optical device for measuring the reflection factor for the perception of a parameter for measuring the study area, the apparatus includes: an optical fiber; source means operable to input optical radiation to the optical fiber probe wavelength; and detection means adapted to generate electrical output signals in response to optical radiation backscattered from the optical fiber; it is characterized in that the optical fiber includes a first segment, which is fed with said optical radiation at the probe wavelength, and a second section located in the test area, said first segment has a higher intensity threshold for non-linear effects than the said second length wherein said source is adapted to supply optical radiation at the probe wavelength in the said first segment with an intensity lower than the intensity threshold of non-linear effects of said first portion, but greater than the threshold intensity of the nonlinear effects of the second segment, the attenuation characteristics of the first segment being such that the intensity optical radiation at the probe wavelength reaching the second section is below the threshold for non-linear effects of said second segment.

According to a first aspect of the present invention also provides a method for optical measurement of the reflection coefficient for determining a parameter measured in the area under consideration, the method includes supplying optical radiation at the probe wavelength into an optical fiber and forming the output electrical signals in response to optical radiation backscattered from the optical fibers, it is characterized in that the optical fiber includes a first segment and a second segment, said second segment has a lower intensity threshold for non-linear effects than the first section, said second section located in the art, optical radiation at the probe wavelength is fed in said first segment with an intensity lower than the intensity threshold of the nonlinear effects of the first segment, but higher than the intensity threshold of the nonlinear effects of the second segment, the attenuation characteristics of said first portion is selected so that the intensity of optical radiation at the probe wavelength reaching the second section, lower than the threshold intensity of said second segment.

According to a second aspect of the present invention there is provided an optical device for measuring the reflection factor for determining the parameter detected in the art, the apparatus includes: an optical fiber; source means operable to supply to an optical fiber of the optical probe radiation with a wavelength having a preselected intensity; and detection means adapted to generate electrical output signals in response to optical radiation backscattered from the optical fiber; it is characterized in that the optical fiber includes a first segment, which is fed with said optical radiation at the probe wavelength, and a second section located in the art, and there is a remote amplifier arranged between the first and second portions operable to compensate for attenuation probe intensity at a wavelength within a first segment of the fiber.

According to a second aspect of the present invention also provides a method for optical measurement of the reflection coefficient for determining a parameter measured in the area under consideration, the method includes supplying optical radiation at a probe wavelength with a preselected intensity into an optical fiber and forming the output electrical signals in response to optical radiation backscatter from the optical fiber is characterized in that the optical fiber includes first and second segments connected by a remote amplifier, the gain of the amplifier is selected to compensate for attenuation at the probe wavelength in the first segment.

According to a third aspect of the present invention there is provided an optical measuring device for determining the reflection coefficient parameter detected in the art, the apparatus includes: an optical fiber; source means operable to supply to an optical fiber of the optical probe radiation with a wavelength; and detection means adapted to generate electrical output signals in response to optical radiation backscattered from the optical fiber; it is characterized in that the optical fiber includes a first segment, which is fed with said optical radiation at the probe wavelength, and a second section located in the area under consideration, said first segment is optically coupled to a second segment at a point remote from the source means, and includes first and second fibers, the first fiber is arranged such that it allows the passage of optical radiation at the probe wavelength towards the second section and the second fiber is positioned such that it allows the passage of optical radiation backscatter return from the second interval to a means of detection.

The following example makes reference to the accompanying drawings, in which:

Figure 1 shows a device according to Oikawa first embodiment of the first aspect of the present invention;

Figure 2 shows a device according to Oikawa first embodiment of the second aspect of the present invention;

3a shows a device according to Oikawa second embodiment of the second aspect of the present invention;

Figure 3b shows in more detail part of the device of Figure 3;

4a shows Oikawa device according to a second embodiment of the second aspect of the present invention;

4b shows Oikawa device according to a third embodiment of the second aspect of the present invention;

4C shows Oikawa device according to a fourth embodiment of the second aspect of the present invention;

Figure 4D shows a device according to Oikawa fifth embodiment of the second aspect of the present invention.

According to an embodiment of the first aspect of the present invention may be proposed a method in which the backscatter power from the remote end of the optical fiber can be increased at the same time allowing that the power supplied to the fiber remains below the limit of nonlinearity. In this method, the optical fiber is made of at least two fiber segments such that the numerical aperture increases monotonically with increasing distance from the measuring device. This method is illustrated in Figure 1, where the measurement device backscattering Oikawa, schematically represented by a source 1, detector 2 and the connector 3, 4 delivers pulses of high power in a first fiber 5. The intensity of pulses 4 is selected below the threshold significant nonlinear effects (depending on the field the use of the system and its design limit the effects can be BPP (stimulated Raman scattering), FBG (stimulated Brillouin scattering), SPM (self phase modulation) or the like.). Cut fiber, which is of primary interest is a second fiber 6 which is connected to fiber 5 at a distance Lf. Fibre 5 is selected so that it has a high threshold for non-linear effects (and therefore, a low backscatter) and the fiber 6 - considerably higher threshold (and thus a significantly higher backscattering coefficient), the relationship between the distance Lf and the effective areas of fibers 5 and 6 are chosen so that the power propagating in fiber 6 is attenuated during its propagation in the fiber 5 to the extent that it is below the threshold of non-linearity in fiber 6, even though it would exceed said threshold with direct its introduction into fiber 6 without passing through fiber 5. Thus the attenuation of fiber 5 is used to bring the intensity of pulses 4 to a level acceptable in the fiber 6 without non-linear effects. Since in certain applications, the fiber 5 is needed to cover a certain distance before reaching the start of fiber 6, the advantage of this method can be substantial.

As an example, the advantages which can be obtained by using this method, the numerical aperture (a quantity which relates to the difference in refractive index between the core and the sheath) fibers extended generally is 0.12 and the corresponding mode field diameter of about 11.5 microns. However, for special applications available fibers having numerical apertures up to 0.29. Since the effective area is inversely proportional to the square of the numerical aperture, and the backscatter factor proportional to its square, it is obvious that for the same supplied power is replaced with a fiber with a higher numerical aperture in the region where the probing power decreases proportionally to a value of at least the square of the numerical aperture of the first and second fibers will result in an increase of the received signal is proportional to the square of said ratio. In the example above, the improvement of the received signal is approximately 6.

Typically, fibers having a large numerical aperture (small effective area), subject to some extent higher transmission losses than fibers having larger effective areas. Therefore, the location of the connection between the fibers 5 and 6 can be optimized. In addition, it is generally desirable that the effective area could be varied in several steps. The reason for changing the properties of the fiber from large to small Aeff in several steps is that splice losses typically lower and, in addition, thereby minimizing the length of the greatest loss (with the lowest Aeff). As an example, the limiting case, the properties of the fibers are gradually reduced so that the losses associated with changes in the properties of the fibers are zero, as changes occur adiabatically and also the ratio of the loss / Aeff along the fiber can be optimized.

Hereinafter, a method for selecting the relative lengths of each section of the fiber consisting of sections of fiber having successively increasing numerical aperture.

With the known ratio between the backscatter level in the i-th and (i + 1) th interval, losses at the interface between them and the attenuation per unit length (i + 1) th segment of the optimal length for (i + 1) th segment It can be obtained from the following reasoning. The shorter the (i + 1) th interval, the higher the backscatter returning from its remote end. However, if this interval is very short, then the signal scattering from the distal end of the previous section will be weaker than the signal from the end of the (i + 1) -th segment. From this it follows that the optimum length of the (i + 1) th segment is given:

<img he="22" wi="65" file="00000003.tif" img-content="undefined" img-format="tif" />

wherein B is absorbed part of the backscatter (i.e., part of the scattered light that is reabsorbed in the waveguide in the reverse direction), αS is the scattering loss for the spectral component used in the measurement, S represents the loss at the interface between section i and section i 1, and α is the total loss in the section. In all cases, the values ​​of which vary from section to section, identified by the index is considered section. In this case, α is expressed in dB / km, S in dB and αS expressed in linear units (as a part extending forward light scattered per unit length). To the resulting optimal length for the (i + 1) th section applies the following assumptions:

1. The resulting length is less than the length of the critical length of interest designed for the highest accuracy, if this condition is not met, then the optimum length is simply the same as that of the end of the segment under consideration.

2. The power remaining in the probe pulse in the compound below the necessary for the emergence of non-linear effect limiting system performance. If this condition is not satisfied, then the connection between the two sections is moved further towards the remote end of the fiber. Moving compound has two effects, namely, to reduce the power of the final segment when it reaches the momentum and also increase in the threshold of the nonlinear effects in the final segment, since the threshold is dependent on the length of the final segment.

When more than two segments of different fibers above procedure is repeated for all the connections between the segments of the fibers, from the most distant.

The methods described above are entirely passive. However, similar advantages can be obtained by using amplification means disposed at a distance from the measuring device. According to a second aspect of the invention, power from a remote fiber segment perceived Oikawa sensor can be increased by employing remote in-line optical amplification means to increase the power level in the final segment of the fiber and to amplify the same amplifier, the signal returning from this segment. Although the description below the amplifier is located in the region and, preferably, the remote pumping amplifier is also possible to remotely fed to the remote location from an electrical source (possibly driving the local pump source).

Remotely-pumped amplifier is a variant of an optical amplifier in which pump power is transmitted in an optical fiber amplifier, and which, therefore, is an electrically passive. Although remote amplifiers are used extensively in telecommunications and arrays of discrete sensors (for example, a pulse poll array of acoustic sensors, US 5,866,898), they have so far not been used in sensors such as Oikawa, especially temperature sensors based on Raman or Brillouin scattering. In the case of Raman scattering, the spectral width of the signal is very high (10-15 THz) which results in a very high noise level at the amplifier input. In the case of the Brillouin scattering spectrum must be filtered backscatter for any benefit from this method. This may be the reason that these methods have not been proposed hitherto.

Oikawa type sensor using a second aspect of the present invention includes at least a first section of fiber into which the pulses are applied, wherein for said first segment should be at least the second fiber segment, the two segments are separated by an optical amplifier. The amplifier preferably includes a length of fiber doped with rare-earth ions, with the property of the gain at the probe and / or signal wavelengths when irradiated by the pumping radiation with a predetermined wavelength. In one preferred embodiment, the signal wavelength is in the range from 1,525 to 1580 nm and amplifier comprises a segment of single-mode fiber doped with erbium ions. Pumping the amplifier it is preferably carried out at a wavelength of approximately 1480 nm and the pumping radiation is preferably transmitted along the same fiber as the probe and backscattered light. The amplifier fiber doped with erbium, particularly suitable for use, meaning that the lifetime of the upper state of erbium ions is sufficient for the device to save significant amounts of energy and thus relatively low pump power can provide efficient amplification for a short period of time (working cycle) for which it is required.

This configuration is illustrated in Figure 2, where the source 1 enters the probe pulses 4 into a first section of fiber 50 through a directional coupler 3. At the end of the fiber segment 5 7 doped with erbium, it is docked in line with the first fiber segment 50 and then the second segment 60 fiber. Second, wavelength selective, coupler 8 is used for the light input 10 of the pump into the fiber, the coupler 8 is selected so that it transmits substantially all sounding signals and backscatter undistorted and enters into the segment 50 of fiber practically all pump radiation which It has a wavelength different from the other signals (probe and the reflected signals). The pumping radiation passes through the fiber segment 50 and partially absorbed by the reinforcing fiber 7. Optionally, may be added wavelength selective reflector 9 at the distal end 7 of the reinforcing fiber for reflecting unabsorbed pump power back into the amplifier, thereby improving the efficiency with which the capacity of the pump used . Reflector 9 is designed in such a way that has minimal impact on the transmission of all signals other than the pump signal. Preferably, it is implemented as a fiber Bragg grating. It should be noted that in this embodiment, the amplifier serves to increase both the level of probe pulses and the signals reflected from the segment 60 of the fiber.

If the gain of the amplifier is chosen to compensate for losses in the first segment 50 of fiber, the system performance would be similar to the system performance, the conductive measuring only a finite segment 60 fibers, but with the following disadvantages: a) the noise figure of the amplifier degrades the signal-to-noise ratio, b) the frequency of repetition pulses limited by the total length of the fiber (sections 50 + 60) and at least limit the scope is determined only by the segment 60, and c) the power which can be introduced into the first section 50 is always less than it would be in the case of measuring only the final segment 60 due to the greater length, which can generate new wavelengths in the probe pulse (the predecessor of the full manifestation of optical nonlinearity).

In a preferred embodiment, the power level of the sensing input to the first segment of the fiber is significantly below the level at which the segment appearing nonlinear effects, and the gain of the amplifier is selected such that its gain exceeds the attenuation in the first segment, and that the power sensing at the output of the amplifier in the second segment is closer to the limit specified for the power of the second segment, defined by restrictions on non-linear optical effects in said second interval.

While intuitively it can be expected that the probe power must preferably be maximized, Oikawa's work type sensor as just described has the following advantages. Certain types of nonlinear effects in optical fibers develop gradually along the fiber. Such a development is retarded with a decrease in the power supplied to the section of the fiber. As a result, power sensing amplifier for fiber lengths can be adjusted to a higher level than would be the case if the power level in the interval before the amplifier would be closer to the limit for that segment. This advantage is illustrated in the following example, which is specific for a single-mode fiber. It is assumed that the restriction on the input power is a result of phase modulation. It is also assumed that the segment under consideration is the last fiber 10 km 30 km fiber amplifier placed immediately before the final 10km, i.e. 20 km from the measuring device. It is also assumed that the maximum acceptable spectral broadening due to SPM is 500 MHz. It is also assumed that the gain of the amplifier is set only for compensation of losses in the first segment of the fiber and that the power input to the first section is maximized. Under these conditions, the power sensing should be limited to a capacity of about 140 MW to 14 nsec Gaussian pulses. In contrast, if the first segment is applied only 50 mW, then the amplifier gain can be selected to deliver 250mW into the final fiber segment with same spectral broadening. Therefore, the development of a decision to restrict capacity in the first segment of the fiber can be increased by 78% sensing capacity in this segment of the fiber. Similar advantages exist when the first limit on the input power is a result of stimulated Raman scattering. For example, if the gain of the remote amplifier is set so that it corresponds to the loss in the first leg of the fiber, it can be calculated that the maximum power that can be supplied, provided that the cumulative loss of nonlinear probe pulse will be lower by 0.3% (which corresponds to an error of 1 ° C for based on the intensity of the Brillouin systems), 0.9 watts. However, the power supplied to the first length of fiber is limited to a level significantly below 0.9 W, for example, 0.1 W, then the amplifier gain can be selected to deliver 2 W into the second section of fiber for the same level of distortion, i.e. increase in power sensing fiber segment in which it matters by a factor greater than two.

A second advantage achieved by the remote amplifier with gain greater than required for simple compensation propagation losses in the first segment of the fiber is applicable to systems where the same amplifier is used for the signal returning from the second fiber segment. In this case, the stronger signal returning from the remote segment of the fiber can be sized to eliminate attenuation of further optical components of the receiving optical fiber lengths for suppressing noise preamplifier stage. Thus, an additional advantage, the corresponding amount (on a logarithmic scale) of optical losses in the measuring device before the next stage plus the noise figure of the gain of this stage may be obtained by reducing the power in the first sensing interval fiber.

In an exemplary embodiment it is assumed that the preferred embodiment is adapted with a fiber amplifier doped with erbium, which is performed remotely-pumped laser at 1480 nm with a sensor in the region of 1550 nm. Assuming that 50 mW is introduced into the first section of 20 km in length and up to 250 mW amplified in an amplifier and, assuming that the propagation loss in the first leg of the fibers of 4.8 dB, we obtain the total required enhancement of 9.8 dB . The average radiation power of 2.1 mW amplifier, in addition to which it is assumed 4nVt power backscattered in the opposite direction of the second segment of fiber. Under such circumstances the segment - about 8 m fiber type NE980 (supplied by Lucent Technologies of New Jersey, USA) will be sufficient to provide the desired gain. Less than 2mW of pump power would be absorbed and with 5mW adopted by the amplifier, the calculated noise ratio is 4 dB. At 1480 nm the fiber losses in the first segment order are assessed as having 0.25-0.3 dB / km, and therefore, will be sufficient to introduce 20 mW in a first section to a power of 5 mW hit. Alternatively, if the input power sensing reduced to 25 mW is sufficient amplification of 15 dB, which can be achieved by means of 12 m length of the same fiber pumped by a 10 mW of optical power at 1480 nm (assuming that the set optional reflector pump ). In the latter example, the signal from the second fiber segment returns to the beginning of the first segment with a value 10 times higher than that with which it would come out of the second segment, having only 4 dB of added noise. Increased signal may significantly reduce the effect of noise amplification at a later stage in the system.

In certain circumstances, it may be required to fulfill certain optical function at a distance, for example, filtering the probe signal before entering the last stretch. The filtering operation eliminates a side effect of light accompanying the probe pulse, such as the broadening due to SPM in the first segment before the critical last leg. In general, these functions can be applied to the main bang or backscattered signal. Additionally, you may need to choose the gain of the amplifier so remote as to be different for the probe pulses and signals returning from the second section. Therefore, it may be necessary to provide an optical assembly located remotely from the measuring device and to a segment of the fiber, which measurement should be performed, which extends frontward (the probe) light is separated from radiation propagating in the reverse direction, and processed before entering the last segment of the fiber. The radiation returning from the last section, can also be processed (eg, amplified) and return to the measuring device. The probe radiation passing from the measuring device assembly may share the same fiber as the light returning from the assembly. Such a construction is shown in Figure 3, where a pair of circulators 13a, 13b is used to split the light at the entrance to the assembly 71, which includes amplifiers 11a and 11b and the filter 12, and for re-combining light before leaving the assembly. The circulator is a particularly effective means of implementing this feature, although it is also possible to combine directional couplers (for the splitting of light) and insulators (to ensure unidirectional propagation) in splitting nodes. Depending on the specifications of cleaving devices can be necessary to use a pump-through couplers which are similar to directional coupler 8 in that their wavelength selectivity allows the pump and other split signals. For clarity this is not shown in Figure 3a, but a more detailed view of the preferred construction assembly 71 is shown in Figure 3b.

3B shows additional couplers 81 a-d. They are designed to remove pump power fiber from the first section 50 immediately prior to the circulator 13a (these devices are generally not suitable for transmission of both pump and signal) and re-entering the pump power amplifiers 11a, 11b, as needed. The pump power is transmitted along the route shown by the arrows, namely into one amplifier 11b (in this case the amplifier for the return signals, again, as a preferred option) and the remaining power is then removed from the signal path and then input to another amplifier 11a. Optionally, filter 12 may include a selectively reflective device which returns the residual pump power back through the amplifiers 11a, 11b to improve the efficiency of the pump.

The sensitivity to losses in the components of the assembly must be carefully considered and it may be necessary carefully to scramble the polarization of the light sent to and / or return of said assembly.

Alternatively, it may be preferable to transmit propagating in a forward direction and power sensing signals returning from the last fiber length of individual fibers. For example, distant optical assembly can be coupled to the measuring device, at least two optical fibers connected to various parts of the measuring device, wherein at least one of these fibers is used for the probe pulse from the measuring device to the assembly and, at least one fiber conducts backward signals returning from the fiber segment, for measuring through the assembly and back into the metering device. One or more additional fibers may be used for pump power to a remote amplifier located within the assembly.

Although this structure requires two fibers in the cable to the remote assembly (and thus adds cost) it eliminates the need for one set of splitting components and allows the remaining combining a very simple device. It also allows the fiber to conducting probe pulse optimized independently of fibers conducting the signal back to the measuring device. Finally, it allows for an increase in the pulse repetition rate to a level that is determined only by the passage of time, the spread in both directions in the last segment of fiber, rather than the entire distance from the measuring device to the remote end. As a result of the measurement in case of removal, followed by 20 km 10 km borehole where measurements should be made only in the well can be reduced by about 3 times.

It is evident that although such a construction has great advantages when used in combination with the previously described aspects of the present invention as shown in Figures 4a-c below, the use of one fiber to transmit the probe pulse to a second segment of the fiber and another fiber to transmit the backscattered signals from second fiber segment to the device detection or the like may be advantageous regardless of whether you use to improve transmission of the second fiber segment means carrying the first or second aspects of the present invention.

4a corresponds to the device of Figure 1, but differs in that the test pulses 4 emitted by source 1 are supplied to the first fiber section 5a. Directional coupler 3 in this case is located in the region immediately before the studied area in which the segment 6 is fiber. Signals returning from the segment 6 fibers pass through coupler 3 and back to the measuring device through a separate fiber segment 5b. According to a first aspect of the present invention, the execution of the segment 6 can be selected with a large coefficient of backscattering than the fiber segment 5a.

4b corresponds to the apparatus of Figure 2, but the same as 4, characterized in that the probe pulses are in the region of interest along the first fiber 50a and returned on a second fiber 50b via the directional coupler 3. The difference from the structure of FIG. 4a is that an amplifying section 7 is connected between the fiber coupler 3 and the segment 60 of the fiber. In this case, coupler 3 is wavelength selective, so that all of the pump power injected into the fiber through a coupler 50a has reached 8 and the coupler 3 is directed into amplifying fiber section 7. 3 branching ratio of the coupler at the wavelength of the signal and probe may be selected so that the discharged unequal fraction of power which, in combination with variable gain remote amplifier 7 can be used to optimize system performance.

The device according to Figure 4c, which corresponds to that of Figure 3, is provided distant optical assembly includes separate parts for the probe pulses and the signals returning from the considered fiber segment 60, which is connected to the measurement device via individual fibers 50a and 50b and with fiber segment 60 through a circulator 13 are not shown in Figure 4C couplers required in the remote assembly 71 for removing the residual power in the amplifier 11a, 50a connected to, and subsequent introduction of the second power amplifier 11b. Optionally can be provided reflecting pump filters for improving the use of the pump power. Coupler 8 may be inserted into fiber section 50b, rather than 50a as shown in Figure 4c.

Alternatively, pump power can be transmitted independently on each of the fibers 50a and 50b, as shown in Figure 4D. In this case, the power of the pump 10 is introduced in the fiber 50a, and the fiber 50d, through separate wavelength-selective couplers 8a and 8b. The relative pump may be adjusted to achieve advances relative power in each of the amplifiers 11a and 11b. In the case of 4d by removing assembly is simplified since no wavelength selective couplers for the pump power between the amplifiers 11a, 11b, which improves the reliability of the assembly. Further improvement can be achieved by replacing the single circulator 13 welded to a conical coupler (not shown). In the case of a welded conical coupler overall efficiency can be very high when establishing an asymmetrical branching ratio such as 90% / 10%, in favor of the signals returning from the segment 60 of the fiber. Although the probe pulses are administered efficiently in such an arrangement, the admissibility of a local amplifier can increase their capacity for compensation and the peak power can be introduced only limited by nonlinear effects in the fiber segment 60.