WO1992010014A1

WO1992010014A1 - A method and apparatus for amplifying an optical signal

Info

Publication number: WO1992010014A1
Application number: PCT/DK1991/000365
Authority: WO
Inventors: Anders Overgaard Bjarklev
Original assignee: Nkt A/S; Lycom A/S
Priority date: 1990-11-30
Filing date: 1991-11-29
Publication date: 1992-06-11
Also published as: AU9017691A; DK285490D0

Abstract

The present invention relates to amplification of optical signals propagating in an optical transmission waveguide. Particularly it relates to a method and apparatus for amplifying optical signal propagating in an optical waveguide using an active optical waveguide characterized in that the active optical waveguide has a longitudinally variable structure in which at least one of the parameters: i) the extent of guiding region, ii) the extent of active dopant, and iii) the refractive index difference, varies along the length of the active optical waveguide in such a way that the ratio F of the extent of active dopant to the extent of the mode field of the pumping light is increasing monotonically in the direction of the pumping light propagation, the values of said ratio F being selected within the range from 0.05 to 1.00.

Description

A METHOD AND APPARATUS FOR AMPLIFYING AN OPTICAL SIGNAL

1. BACKGROUND OF THE INVENTION

Field of the invention

The present invention relates to amplification of optical signals propagating in an optical transmission waveguide. Particularly, it relates to a method and an apparatus for amplifying optical signals propagating in an optical wa¬ veguides. More particularly it relates to an Er-doped ac¬ tive optical waveguide, a method of its manufacture, an optical amplifier comprising such a waveguide, and an op- tical communication system comprising such an amplifier.

Active waveguides according to the invention are particu¬ larly useful as light sources, optical preamplifiers, in-line amplifiers, distributed amplifiers, booster am- plifiers, and sensors, particularly distributed tempera¬ ture sensors.

Amplification by use of fiber-optic amplifiers is very usefull in a wide range of communication systems which are not dispersion limited because the optical fiber at¬ tenuation can be compensated in this way. Unfortunately, the fiber-optic amplifiers also contributes to the noise in such systems. This puts some demands on the system design.

In fiber amplifiers, the most important problem is the noise generated. The noise arises as the active fibers emits both spontaneous photons and amplified signal photons. The spontaneous-spontaneous beat noise within the receiver band width and the signal-spontaneous beat noise signals give rise to a noise penalty in a communi¬ cation system with optical fiber amplifiers. One of the objects of the present invention is to reduce this noise level.

Another problem in the design of a practical fiber ampli¬ fier is the optimization of the fiber length. This fiber length depends i.a. on the signal wavelength, the pump wavelength and the pump power for a given fiber. As one of the objects of the present invention this dependency- is reduced.

Still another problem in the design of a practical fiber amplifier is reflections from connectors, components or splices. Such reflections will together with the high gain of the doped fibers, give rise to lasing - an effect which is certainly- unwanted. Still another object of the present invention minimises the risk for such reflec¬ tions.

Finally, losses which arise at the coupling between the amplifier and the rest of the system is a problem because it has to be compensated by a higher gain. The present invention reduces these coupling losses.

Prior art disclosure

EP Patent Application No. 0 180 861 discloses an in-line amplifier of a selected relatively broad bandwidth of light energy having a central wavelength of approximately 1.34 mn which amplifier is based on a gain media wave¬ guide consisting of Nd-doped glass pumped by a laser diode selected to have a light output at a wavelength within the range 0.78 - 0.88 urn. Noise developed as a result either of the pumping light of a laser diode or by the gain medium, i.e. in the form of wavelenghts outside of the carrier bandwidth, is isolated in the gain medium fiber and kept from passage by a wavelength selective, dispersive coupling. The use of such wavelenght dependent couplers or the use of optical bandpass filters mainly reduces the noise problem, but does not result in a lower spontaneous emission for a given gain and input signal power band.

EP Patent Application No. 0 313 209 discloses an optical fiber preferably a single mode fiber having a fluorescent rare earth element additive, e.g. Er or Nd, unevenly dis- tributed over the cross section of the core and having a higher concentration of the additive at the center of the core than at the core/cladding boundary. This confined core-type optical fiber ideally provides an efficient use of pump power launched into the optical fiber, i.e. high gains are achieved for small launch powers, and reduces the problem of noise. However, there are a number of disadvantages.

One problem is related to the manufacture of a confined core optical fiber by the MCVD process in which dopant is evaporated from the center core region during the collap¬ se of preform. Thus, even if Al can be used to reduced the evaporation of dopant, the amount of rare earth ele¬ ment additive is sensitive to the preform collapse.

Another problem is the possibility of diffusion of dopant out of the inner core to the outer core or cladding where it may contribute to the noise. Further, it is difficult to match the spot size of a standard fiber to a high gain confined core fiber having a high refractive index difference, preferably more than 0.03, and being single moded at the pump wavelength at 980 nm or 1480 nm, which requires a core radius smaller than 2.5 urn.

Other problems involving sensitivity to pumping condi¬ tions and reflections and the high coupling losses are not solved by using this confined core technique. Furthermore efficient confined core fibers are rather difficult to produce in a reproducible way if the con¬ f nement is high.

International Application No. PCT/JP90/01499 (Int. Publi¬ cation No. WO 91/07689) discloses an optical amplifier comprising an optical fiber doped with rare earth ele¬ ments of which the diameter of the doped region decreases gradually in the direction of the propagation of the pum- ping light, whereby an enhanced amplification efficiency, defined as utilization of pumping light power, is claimed to be obtained. Nothing is indicated or suggested about the noise reduction problem.

US Patent No. 3 617 917 discloses a solid state laser element comprising a transparent body of host material including a laser-active material substantially uniformly distributed in at least a portion of its entire length wherein the radial refractive index profile of the host material, i.e. either the refractive index at the axis, the shape, or both, decreases monotonically along the axis from the signal input end to the signal output end, thereby increasing the spot size of the signal light beam at the output end and avoiding saturation phenomenon in the vicinity of the laser output end surface, whereby a relatively larger power output is ensured. Nothing is indicated or suggested about reduction of noise.

2. SUMMARY OF THE INVENTION

Objects

In a primary aspect, the object of the invention is to provide an improved method and apparatus for amplifying a light signal propagating in an optical transmission wave¬ guide in which a relatively high gain and a reduced noi¬ se is obtained. Further, it is the object of the present invention to provide an improved method and apparatus for amplifying a light signal propagating in an optical transmission wave¬ guide in which coupling losses and reflections are redu¬ ced, the spectral gain profile is improved, and which amplification and receiver sensitivity have a reduced dependency of the length of the active optical waveguide.

(a) Method of Amplifying a Light Signal

According to the invention these objects can be achieved by providing a method of amplifying a light signal propa¬ gating in an optical transmission waveguide comprising the steps defined in the preamble of claim 1

wherein the active optical waveguide has a longitudinally variable structure in which at least one of the parame¬ ters:

i) the extent of guiding region; ii) the extent of active dopant; and iii) the refractive index difference,

varies along the length of the active optical waveguide in such a way that the ratio F of the extent of active dopant to the extent of the mode field of the signal light is increasing monotonically along the direction of the pumping light propagation, the values of said ratio F being selected within the range from 0.05 to 1.00.

It turns out that noise of an active waveguide is mainly determined by the population inversion of active dopants at the signal input end of the active waveguide, i.e. the end where the signal is introduced. Generally, the noise generation is low in regions exhibiting high population inversion, which are normally present at the center of the active waveguide, but it is higher in regions exhibi- ting low population inversion, such regions being mainly present at off-center regions of the active waveguide at the input end of the signal.

Therefore, according to the invention, by avoiding or reducing the low population inversion of active dopants located off the center in the signal input end of the active waveguide, a relatively high gain and low noise of the amplified signal can be obtained at the output com- pared to active waveguides having a low population inver¬ sion of dopants located off the center through out the length of the waveguide.

It is assumed that due to the transverse distribution of the waveguide modes, in which the guiding modes are con¬ centrated at the center and diluted away from the center of the guiding region, noise will mainly be generated due to spontaneous emission of light from the active dopants in regions away from the center. In these off-center regions, there is a low population inversion of active dopants. Consequently, spontaneous emission in these regions generate a poor signal-to-noise ratio.

According to the invention, a controlled population in- version of active dopants in an active waveguide can be obtained by providing an active optical waveguide having the above defined longitudinally variable structure, in which the ratio F of the extent of active dopant to the extent of the mode field of the signal light is increased monotonically along the direction of the pumping light propagation. Thereby it is ensured that the low popu¬ lation inversion of active dopants off the center of the active waveguide is avoided or reduced and consequently the noise is reduced.

By the extent of active dopant is meant the extent of the guiding region doped with active dopant. According to preferred embodiments:

the light signal is introduced into the active optical waveguide co-propagating with the pump light;

the light signal is introduced into the active optical waveguide counter-propagating with the pumping light;

the extent of guiding region and the extent of active dopant is increasing monotonically from the end where the pump light is introduced to the other end and the re¬ fractive index difference is constant;

the extent of guiding region and the extent of active dopant are constant and the refractive index difference is increasing monotonically from the end where the pump light is introduced to the other end; and

the extent of guiding region and the refractive index difference are constant and the extent of active dopant is increasing monotonically from the end where the pump light is introduced to the other end.

According to the invention, the active dopant is a fluorescent additive, particularly a rare earth element selected from the group consisting of Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu, preferably Er, Nd, Pr, Ho, Yb, and Tm, most preferably Er.

According to the invention, the active optical waveguide includes the active dopant in a concentration from 5*10 19

_ _ — -\ to 1*10 ions m , preferably for distributed amplifiers from 1*10 20 to 10*1020 ions m-3, and preferably for short amplifiers from 1*10 to 2*10 ions m^~ . "Transverse Mode Field of the Signal Light"

According to the invention the transverse mode field of the signal light can be varied by varying the extent of the guiding region, the refractive index difference between the guiding region and the cladding, or both.

Thus, for an active optical fiber waveguide e.g.. having a core diameter less than 4 j_m and a refractive index dif- ference larger than 0.015, the transverse extent of the mode field of the signal light can be increased by de¬ creasing the diameter of the core, or by decreasing the refractive index difference.

Also, the transverse extent of the mode field can be changed by changing the difference of refractive index between the guiding region and cladding of the waveguide. A large difference of refractive index implies a small transverse extent of the mode field of the signal light an vice versa.

"Transverse Extent of Dopant"

According to the invention, the active dopant can be ar- ranged in the guiding region with a variable or constant transverse extent. Thus, the active dopant is confined in the guiding region, wholly or partially filling up the guiding region.

In a preferred embodiment, the active dopant is arranged in the guiding region to the full transverse extent of the guiding region.

(b) Active Optical Waveguides

According to a secondary aspect, it the object of the invention to provide an active optical waveguide which has regions of reduced low population inversion.

According to the invention this is obtained by providing an active optical waveguide suitable for amplifying a light signal having a predetermined wavelength, com¬ prising:

i) a guiding region (12,22) containing an active dopant arranged in a region (121, 221) thereof; and

ii) a cladding (11,21) surrounding the guiding region and determining the extent thereof, and having a refractive index less than the refractive index of the guiding region,

whereby the mode fields of τhe propagating light signal and the pump light are determined by the extent of the guiding region, the refractive index difference between the guiding region and the cladding, or both,

i) the extent of guiding region; ii) the extent of active dopant; iii) the refractive index difference,

is varied from a first end of the active optical wavegui¬ de to a second end in such a way that the ratio F of the extent of active dopant to the extent of the mode field of the signal light is increased monotonically from the first end to the second end, the values of said ratio F being selected within the range from 0.05 to 1.00. Generally, the F ratio is selected within the range from 0.05 to 1.00, preferably from 0.1 to 0.8, most preferably from 0.25 to 0.8.

According to preferred embodiments:

the extent of guiding region and the extent of active dopant is increasing monotonically from the end where the pump light is introduced to the other end and the re- fractive index difference is constant;

According to the invention, the active optical waveguide includes the active dopant m a concentration from 5*10 19 to 1*10 ions π_~ , preferably for distributed amplifiers from 1*10 20 to 10*1020 ions m-3, and preferably for short oo C _o amplifiers from 1^•10 to 2^•10 ions m

In a preferred embodiment, the active waveguide is an ac- tive optical fiber waveguide, particularly a tapered ac¬ tive optical fiber waveguide. 11

( c) Methods of Manufacturing a Tapered Active Optical Fiber

According to another aspect, it is the object of the pre¬ sent invention to provide a method of manufacturing an active optical fiber waveguide exhibiting a relatively high gain and a reduced noise.

10

These objects can be obtained by a method of manufac¬ turing an active optical fiber waveguide which is characterized in

15 a) providing a preform having a longitudinally structure in which

i) the transverse extent of the core, the transverse extent of dopant, and the refractive index dif- 20 ference are constant;

ii) the transverse extent of the core is varied, and the transverse extent of dopant and the refractive index difference are constant;

25 iii) the transverse extent of the core and the transverse extent of dopant are constant, and the refractive index difference is varied;

30 iv) the transverse extent of the core is constant, the transverse extent of dopant is varied, and the refractive index difference is constant; and

v) all three parameters are varied, and

35 b) drawing the preform to the active optical fiber waveguide, whereby a tapered form is drawn in case ( 1 ) .

"Constant Parameter - Tapered Fiber"

In a preferred embodiment there is provided a method of manufacturing an active optical waveguide suitable for amplifying a light signal having a predetermined wave¬ length, which is characterized in

a) providing a preform having a longitudinally constant structure with parameter values selected within the fol¬ lowing ranges:

i) a constant transverse extent of the core providing a final core radius from 0.5 to 20 um of the active optical fiber waveguide;

ii) a constant transverse extent of dopant in the full extent of the core;

iii) a constant refractive index difference between the

-3 core aanncd cladding in the range from 5*10 to

b) drawing the preform into the active optical fiber waveguide having the outer diameter and the core diameter of the fiber changing monotonically with a factor of 1.1 - 4 over a length from 1 m to 1000 m; and

c) selecting the values of the above mentioned parameters to ensure that the values of the ratio F of the extent of active dopant to the extent of the mode field of the signal light fall within the range from^' 0.05 to 1.00. "Variable Core Extent- Straight Fiber"

In a preferred embodiment there is provided a method of manufacturing an active optical waveguide suitable for 5 amplifying a light signal having a predetermined wavelength, which is characterized in

a) providing a preform having a longitudinally variable structure with parameter values selected within the fol- 10 lowing ranges:

i) a variable transverse extent of the core providing a final core diameter from 0.5 to 20 um of the active optical fiber waveguide, said extent of the final 15 core varying a factor of 1.1 to 4 from the one end of the final active optical fiber waveguide to the other;

ii) a constant transverse extent of dopant in the range 20 from 0.05 to 1.00 times the core diameter;

iii) a constant refractive index difference between the

_3 core aanncd cladding in the range from 5*10 to

25 b) drawing the preform into the active optical fiber waveguide with a length of 1 - 100 km; and

c) selecting the values of the above mentioned parameters 30 to ensure that the values of the ratio F of the extent of active dopant to the extent of the mode field of the signal light fall within the range from 0.05 to 1.00.

"Variable Index Difference - Straight Fiber"

35

In a preferred embodiment there is provided a method of manufacturing an active optical waveguide suitable for amplifying a light signal having a predetermined wavelength, which is characterized in

a) providing a preform having a longitudinally variable structure with parameter values selected within the fol¬ lowing ranges:

i) a constant transverse extent of the core in the ran¬ ge providing a final core radius from 0.5 to 20 nm of the active optical fiber waveguide;

ii) a constant transverse extent of dopant in the range from 0.05 to 1.00 times the core diameter;

iii) a variable refractive index difference between the core and cladding varying a factor of 1.1 to 4 from the one end of the preform to the other in the range from 5*10^"3 to 50*10~³;

t>) drawing the preform into the active optical fiber waveguide with a length of 1 - 100 km; and

c) selecting the values of the above mentioned parameters to ensure that the values of the ratio F of the extent of active dopant to the extent of the mode field of the signal light fall within the range from 0.05 to 1.00.

"Variable Dopant Extent - Straight Fiber"

a) providing a preform having a longitudinally variable . structure with parameter values selected within the fol¬ lowing ranges: i) a constant transverse extent of the core providing a final core diameter from 0.5 to 20 am of the active optical fiber waveguide;

ii) a variable transverse extent of dopant varying a factor of 1.1 to 4 from the one end of the preform to the other in the range from 0.05 to 1.00 times the core diameter;

iϋ) a constant refractive index difference between the

_3 core ainncd cladding in the range from 5*10 to

50*10 ^■3

b) drawing the preform into the active optical fiber waveguide with a length of 1 - 100 km; and

"Variable Parameter Fiber"

i) a variable transverse extent of the core providing a final core diameter from 0.5 to 20 .m of the active optical fiber waveguide, said extent of the final core varying a factor of 1.1 to 4 from the one end of the final active optical fiber waveguide to the other;

iii) a variable refractive index difference between the core and cladding varying a factor of 1.1 to 4 from the one end of the preform to the other in the range from 5-10^-3 to 50*10~³;

"Transverse Extent of the Guiding Region"

A change of the extent of the core region can be obtained by drawing a preform to a fiber so that the outer dia- meter of the fiber and thereby the core diameter are changed during the draw.

A change of the extent of the of the core region can be obtained by increasing or decreasing the diameter of the core of the preform.

'Transverse Extent of Dopant"

The extent of the doped region can be varied by changing the extent of a guiding region fully confined with active dopan . The transverse extent of dopant can be varied by varying the number of layers of dopant during the production of the preform. Thus, if the mode field of a fiber drawn from a preform produced by the MCVD process is to be kept constant and the extent of dopant varied, the core of the preform can be deposited in several steps, each of which provides approximately the same refractive index diffe¬ rence between the core and the cladding by depositing layers having different decreasing or increasing thick- nesses, respectively. Thus, by varying the thickness of the last layer containing dopant, the extent of dopant is varied at a constant index of refraction, i.e. at a constant mode field.

"Rejf ac ive Index Difference "

The index of refraction between the guiding region and the cladding is provided according to methods known in the art.

Thus, dopants like Ge and Al which change the refractive index are used to increase the refractive index.

Primarily, Al is used to reducing the loss of rare earth element dopants during prefrom collapse. Further, F de¬ creases the refractive index.

In the case of an Er-doped optical fiber waveguide am¬ plifier intented for a distributed amplifier having a length from 1 km up to 100 km, or more, varying the structure of the guiding region of the preform.

In case of an Er-doped optical fiber waveguide with a high concentration, the tapering is performed from about 1 to about 1000 m, or more, during the draw. Tapering of an optical fiber waveguide is performed by continu¬ ously changing the outer diameter of the fiber, e.g. 1 wu/m.

Tapering according to the invention should be distingui¬ shed from tapering of drawn optical fibers for the pur- pose of mode-coupling of optical fibers having different mode field diameters. In this case the tapering is per¬ formed over a length of no more than a few millimeters.

(d) Optical Amplifier

According to still another aspect, it is the object of the invention to provide an optical amplifier which has a high gain and a reduced noise.

According to the invention, this is obtained by an opti¬ cal amplifier comprising:

a) a source of light signals (10); b) a source of pumping light (20); c) an active optical waveguide (30);

d) coupling means for coupling pumping light into an end of the active optical waveguide;

e) coupling means for coupling the light signals into an end of the active optical waveguide; and

) coupling means for coupling the amplified light signals out of the active optical waveguide;

wherein the active optical waveguide has a longitudinally variable structure in which at least one of the parame- ters: i) the extent of guiding region; ii) the extent of active dopant; iii) the refractive index difference,

is varied from a first end of the active optical wavegui¬ de to a second end in such a way that the ratio F of the extent of active dopant to the extent of the mode field of the signal light is increased monotonically from the first end to the second end, the values of said ratio F being selected within the range from 0.05 to 1.00.

Generally, the F ratio is selected within the range from 0.05 to 1.00, preferably from 0.1 to 0.8, most preferably from 0.25 to 0.8.

According to preferred embodiments:

According to the invention, the active dopant is a fluorescent additive, particularly a rare earth element selected from the group consisting of Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu, preferably Er, Nd, Pr, Ho, Yb, and Tm, most preferably Er. According to the invention, the active optical waveguide includes the active dopant in a concentration from 5*10 19 to 1*10 ions m , preferably for distributed amplifiers from 1*10 20 to 10-1020 ions m-3, and preferably for short oo c ___o amplifiers from 1*10 to 2-10 ions m

In a preferred embodiment, the active waveguide is an ac¬ tive optical fiber waveguide, particularly a tapered ac¬ tive optical fiber waveguide.

Further, it is preferred that the transverse mode field of the active optical waveguide is matched to the tran- verse mode field of a fiber transmiting the source light signals.

In another preferred embodiment, the the transverse mode field of the coupling means is matched to the transverse mode field of the active optical waveguide.

in still another embodiment, the coupling means comprises wavelength selective means.

Further, according to another preferred embodiment an optical amplifier is provided which is characterized in that it further comprises a second active optical wave¬ guide, the second end of which is connected to the second end of the first active optical waveguide, and the first end of the second active optical waveguide is connected to the coupling means for coupling the amplified light signals out of the second active optical waveguide, and that the pumping light source is coupled to the first end of the active waveguide.

For system applications, semiconductor lasers are the preferred pump laser based on lifetime considerations. Further, coupling means to be used are known in the art, e.g. fused bicon taper couplers (WO 83/04409), couplers using dichroic beam splitters, couplers using gratings, and couplers using curved side-polished fibers (WO 84/02005). The fused coupler or the side-polished coupler is preferred when low coupling loss is required. When wavelength selection is required, the side-polish type, the dichroic interference filter, or the diffration grating type is preferred. Particularly, if high wave- length selectivity is required with a 1480 nm pump, the interference filter or the diffraction grating type is preferre .

(e) Fiber optical transmission system

According to still another aspect, it is the object of the invention to provide an optical transmission system which has a low noise and low coupling losses and reflections.

According to the invention, this is obtained by providing an optical transmission system, comprising:

a) a transmitter (60); b) a receiver (70); c) an optical fiber (80); and d) one or more amplifiers or regenerators (90);

wherein one or more of the amplifiers are amplifiers ac¬ cording to the invention.

Definition of Terms

Within the context of the invention, the term "active optical waveguide" is intended to designate an optical waveguide comprising an active gain or lasing medium which absorbs light energy at one or more absorbing wavelengths and emits lights at a different emitting wavelength or at more different wavelengths.

Further, within the context of the invention, the term "mode field" is intended to designate the Laplace mode field which can be measured by Far Field Scan Techniques (see e.g. Recommendations from CCITT G.652, "Characteris- tics of Single-Mode Optical Fiber Cable", Annex A and B, Melbourne 1988). The extent of mode field is also termed "spot size".

3. BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1A illustrates the principle of a longitudinal sec¬ tional view of an active optical fiber according to the prior art;

Fig. IB shows a cross sectional view along the lines A-A and B-B in Fig.l; and

Fig. 1C illustrates the refractive index difference and the mode field along the lines A-A and B-B in Fig.l;

Fig. 1D-1F shows a longitudinal sectional view of an ac¬ tive optical fiber according to the invention;

Fig. 2A-2C shows a preferred embodiment of a tapered ac¬ tive optical fiber;

Fig. 3A-3C shows another preferred embodiment having a variable extent of the guiding region and of the dopant at a constant refractive index difference; Fig. 4A-4B shows another preferred embodiment having a constant extent of the guiding region and of the dopant at a variable refractive index difference;

Fig. 5A-5C shows another preferred embodiment having a constant extent of the guiding region and a variable extent of the dopant at a constant refractive index difference;

Fig. 6 illustrates an embodiment of an optical amplifier according to the invention;

Fig. 7 illustrates a fiber optical transmission system;

Fig. 8 shows the receiver sensitivity versus length of active fiber for a preamplifier based on an untapered active optical fiber waveguide and a tapered active op¬ tical waveguide according to the invention;

Fig. 9 shows the - 40 dBm receiver sensitivity limit as a function of pump power and fiber length for a preampli¬ fier based on an untapered active optical fiber waveguide and a tapered active optical waveguide according to the invention;

Fig. 10 shows the gain versus fiber length for a pream¬ plifier based on an untapered active optical fiber wave¬ guide and a tapered active optical waveguide according to the invention;

Fig. 11A-11B show gain and noise figur versus fiber length for a preamplifier based on an untapered active optical fiber waveguide and a tapered active optical waveguide according to the invention;

Fig. 12 shows pump spotsizes versus core radius for different index differences; Fig. 13 shows signal spotsizes versus core radius for different index differences;

Fig. 14-16 show doping radius-spotsize ratios versus cladding radius, index difference, and confinement fac¬ tor, respectively;

Fig. 17 illustrates doping radius-spotsize ratios versus fiber position coordinate for a "thin", a "thick", and an uptapered fiber;

Fig. 18 diagrammatically illustrates a test set-up for testing an active optical waveguide;

Fig. 19 shows gain versus pump power of tested tapered and straight active optical fibres of type 1;

Fig. 20 shows noise figure versus pump power of the tapered and straight fibres in Fig. 19;

Fig. 21-22 show the similar gain and noise figures of the tested type 2 fibres; and

Fig. 23 and 24 show calculated gain and noise figure, respectively, of uptapered and straight active optical ibres.

4. DETAILED DESCRIPTION

Active Optical Waveguides

Fig. 1 describes the general principle function of an active optical waveguide according to prior art and the invention.

Referring to Fig. 1A, an active optical waveguide accord¬ ing to prior art comprises a guiding region 12 fully do- ped with dopant and a cladding 11. Therefore, this fiber has acon=1. Further, the fiber has a step-index p^rrofile,

Fig. IB shows cross sections at the lines A-A and B-B.

Fig. 1C shows the pump light intensity profile described by the mode fields 122 and 222 at the lines A-A and B-B, respectively, and the refractive index profile of the guiding region 121 and 221 at the lines A-A and B-B, re- spectively.

When pumped with pump light at one pump end, regions of low population inversion 13 in the opposite end of the waveguide are present in the outer region of the doped core 12.

According to the invention, these regions of low popula¬ tion inversion 13 can be reduced by providing a longi¬ tudinally variable structure of the active optical wave- guide as illustrated in Figs. ID, IE and IF.

Fig. ID shows the pumping configuration with a counter- directional pump and signal for definition of the points were the pump intensity profiles have been evaluated. If the pump is introduced into the active optical waveguide in the thin end (A-A), the pump spotsizes will be the same. Pumping in the thin end is the preferred embodiment of this invention.

In the Fig. IE and IF, the pump light intensity spotsizes are shown together with the dopant profile which in this example is identical with the index profile. It can be seen that the pump spotsize radius is significantly lar¬ ger than the dopant radius in the thin end of the fiber which is the pump input end according to the present in¬ vention. The Fig. 2,3,4 and 5 shows preferred embodiments.

In Fig. 2 and 3, the extent of guiding region 12 in the one end of the active waveguide is fully doped with dopant, and thus the extent of dopant is increased to a higher value at the other end 22 at a constant index of refraction 121, 221.

Fig. 2 illustrates one first preferred embodiment of the invention where a preform with a constant outer diameter, a constant core diameter, a constant index difference, and a constant dopant confinement has been produced. This preform has been drawn with a varying drawing speed to obtain a linearly tapered fiber as can be seen in Fig. 2A. The pump light is to be introduced in the thin end

(A) of the fiber. The signal light can be introduced both in the thin end or in the thick end.

Fig. 2B and 2C show the pump modefield distribution 122 and 222 together with the active doping region 121 and 221 which follows the refractive index profile for this fiber.

As the noise generation is mainly defined by the popula- ion inversion in the beginning of the fiber, this tapered fiber with the low dopant radius/pump spotsize radius at the input end of the light signal will have a lower noise figure than the straight fiber of Fig. 1A-1C.

Fig. 3 illustrates another preferred embodiment. The preform is produced with a constant outer diameter but with a varying core diameter. The refractive index dif¬ ference is constant and the active dopant confinement is also constant. The preform can be manufactured by gra- dually changing the burner traversing speed when the core is deposited in the MCVD process. If the fiber is drawn with a constant outer radius, the core radius and the active dopant radius will change.

This fiber design only differs from the design in Fig. 2 by the cladding radius. Consequently, the noise genera- tion as compared to a straight fiber will be lower with¬ out decrease of gain in the fiber.

In Fig. 4, the extent of the guiding region and the extent of dopant is constant, but the refractive index is varied from 121 at the one end to 221 at the other.

Such a fiber can be produced by changing the refractive index raising dopant concentration by increasing the flow or e.g. GeCl, in the MCVD proces.

In Fig. 4B, the pump spotsize is shown together with the index profile. The active region is constant in radius and concentration whereas the core is increased in re¬ fractive index from A to B.

The pump is to be introduced in the end A like in the above preferred embodiments. This will result in a low dopant radius to pump spotsize ratio and thereby to low noise generation and high gain.

In Fig. 5A the extent of dopant is variably confined within the extent of the guiding region. Thus, the variable extent of dopant 15 at the one end increases to the full extent of guiding region.

This varying confinement can be realised by controlling the time that different parts of the preform is submerged into the solution of the active dopant ions in the SODOF process.

Fig. 5B shows the active dopant region 15 compared to the core 12 at the pump input end of the fiber and at the output end where the confinement factor is 1 in this example.

The pump spotsize is constant as the waveguide has not been changed, but as the dopant radius is increased from the pump input end to the output end, the ratio between dopant radius and pump spotsize radius will also change. The pump input end is the end where the confinement factor acon is smallest,

Fig. 6 shows schematically a typical amplifier con¬ figuration with a signal source 10, a pumping source 20, a power combiner 40, typically a fiber coupler, an active tapered waveguide 30 according to the present invention and output coupling means 50. This amplifier works in the co-directional configuration with the pumping light and the signal light propagating in the same direction. Thus, the active waveguide 30 has to have the end with a low dopant radius to pump spotsize radius connected to the power combiner 40. Other pumping configurations with counter propagating pump and signal are possible. In all configurations, the end with the low dopant radius to pump spotsize radius should be connected to the power combiner in order to obtain low noise generation.

Fig. 7 shows a typical optical transmission systsem with a transmitter 60 and a receiver 70. optical amplifiers can be introduced in such systems in order to compensate for loss in the waveguide 80 or in the optical power splitters in the system.

The amplifier 90 can be placed at different places in such systems such as just at the transmitter, the so- called booster amplifier, or just before the receiver, the so-called preamplifier. As the signal at the input of the amplifiers placed at different positions in the system is different, the amplifiers will have different optimal constructions, e.g. different fiber lengths. The amplifiers disclosed in this invention are preferably used as preamplifiers, but they can also be used at other positions in the system.

"Doping Radius-Spotsize Ratio, F"

For illustration, we define the extent of active dopant to the extent of modefield F for a cylindrical active waveguide in the following way:

F = doping radius / pump modefield radius

The doping radius is coupled to the basic waveguide para¬ meters at a given point of the fiber as soon as the pre- form has been collapsed. The basic fiber parameters are:

* core radius, ' rcore

* cladding radius, r_cladding,

thus we get at a given point in a fiber:

cladding c core'

where kc >>1.

Normally the active doping radius is equal to or smaller than the core radius:

ract.i.ve = acon*rcore

where acon is the active doping - confinement factor and 0

< acon < 1. We now find the F given by the cladding radius and the modefield radius:

F = (rcl,ad.d.i.ng*acon)"/(kc*w) '

The modefield or spotsize radius is more complex. Follow¬ ing Marcuse [D. Marcuse: "Loss analysis of singlemode fiber splices", Bell Systems Tech. Journal 3, Vol. 56, 1977, pp. 703-718], we find:

w/ 'rcore = 0.65 + 1.619*V^~3/2 + 2.879*V^-6,'

where V is defined

27T

2 2 V = ^r _COre ^^{n (core)} " ^{n (}cladding⁾] λ

27T rcore V[2nΔn]^J .

It should be mentioned that this expression is only approximative, and it thus only describes the modefield qualitatively. Moreover, the V-value is of course given by the precise index variation, wavelength, etc. , and not the approximative ste -index, etc.

If some of the following waveguide parameters are varied along the fiber:

* Core radius

* Refractive index difference between core and cladding

the V-value will vary, and consequently, the modefield radius will vary. In figure 12, the pump spotsizes are shown for different index differences as a function of core radius. The pump wavelength is 980 nm. In figure 13, the corresponding variation in the signal spotsize has been calculated for a signal wavelength of 1550 nm.

There is therefore not a very simple equation, which can be set up for F.

(a) Variation of Core Radius

For illustartion, consider an active optical fiber pro¬ duced from a preform with a fixed k , an index difference of 0.02 and a fixed confinement of acon = 1. The fiber is produced using the standard technique described below.

The fiber is uptapered from an outer diameter of 85 nm to 125 nm by pulling the fiber at varying pulling speed, the F factor can be calculated to follow a curve like in figure 14. It can be seen that the F factor is increased gradually along the length of the uptapered fiber.

(b) Variation of Index Difference

For illustration, consider a straight active optical fiber produced from a preform with a fixed kv__, a fixed confinement of acon of 1, a fixed active dop ^cant con- centration Cacxii.ve and an index difference, which is increased from the input end to the output end. This fiber is produced by changing the burner traversing speed when the soot layers for the core and active ions are deposited.

The fiber is pulled to obtain a straight fiber, and the index difference is changed from 0.01 to 0.02 over the length of fiber considered. This will result in a varia- tion of F along the fiber like shown in figure 15. (c) Variation of Confinement Factor

For illustration, consider a straight active optical fiber produced from a preform with a fixed kc, a fixed index difference of 0.02 and a concentration of the active dopant Cact.i.ve _ and a confinement acon, which is increased from the input end to the output end. The fiber is produced by changing the time the preform is inserted in the active dopant solution by lifting it up slowly from the solution.

The fiber is pulled to obtain a constant outer diameter, and the confinement is changed from 0.5 to 1.0. If the concentration C .. is constant, this will lead to a variation of F like in Fig. 16. If the concentration also changes from e.g. 10 ppm to 20 ppm the result will be like in figure 16.

"Gain and Noise Performance"

The gain and noise performance of doped fibers is coupled to the fiber design. This is illustrated in the following way.

Consider three doped fibers 1,2 and 3 having the doping radius-spotsize ratios as indicated in figure 17. All fibers are drawn from the same preform with an index difference of 0.01, which leads to a numerical aperture of 0.17, if a step index profile is assumed (see for example Luc B. Jeunhomme: "Singlemode Fiber Optices", Optical Engineering series, Marcel Dekker, 1983).

The LP.₁ cut off wavelength λ can be calculated from

(see Jeunhomme, ibid): 2τr λco = rcore V[2nΔn]

2.401 If the cut off wavelength at the output end is 910 nm, this results in a cut-off of 620 nm at the input end. From the correlation between gain and cut off wavelength, the following gain coefficients can be derived by methods known in the art:

620 nm cut-off fiber 2.1 dB/mW 910 nm cut-off fiber 2.8 dB/mW

The gain of the "thick" fiber, fiber 2, with cut-off of 910 nm, is therefore significantly higher than the "thin" fiber, fiber 1. By comparison between a straight fiber and an uptapered fiber, fiber 3, it is seen that the gain (2.8 dB/mW) of a straight fiber with the high cut-off and an uptapered fiber having a gain of 2.8 T 0.5 dB/mW is almost the same (within 0.5 dB).

The noise figure of an amplifier is mainly given by the population inversion in the end of the fiber, where the signal is introduced. A high inversion will lead to a low noise figure. That is the reason for the generally lower noise figures of the copropagating pump configurations as compared to the counter-propagating configurations.

If the active doping is placed in a narrow region in the center of the fiber and if this region is narrower than the spotsize of the pump, low noise generation is obtain¬ ed because all volumes of actively doped core have a high pump intensity and thereby a high inversion. On the other hand the opposite situation will result in doped volumes with low pump intensity and thereby low inversion and high noise generation.

The F value is very convenient for describing the noise generation as this characteristic parameter is showing the ratio between the doping radius and the pump spot- size. A low F value will therefore give a low noise generation.

Thus, we will find a low noise figure for fiber 1 and fiber 3 as the signal input end has a low F value. Fiber 2 shows a higher noise generation as F is higher.

The comparison between the three fibers therefore gives:

Fiber Gain NF

FaJricatiσn Techniques

Fabrication techniques for making preforms and fibers doped with rare earths according to the invention are well known to those skilled in the art. Relevant disclosures include:

a) Vapor-phase techniques:

S.B. poole, et al. "Fabrication of low-Loss Optical Fi¬ bres Containing Rare-Earth Ions", Electron. Lett., Vol. 21, pp. 737-738, 1985;

S.B. Poole, et al., "Fabrication and Characterization of Low-Loss Optical Fibers Containing Rare-Earth Ions", J. of Lightwave. Technol., Vol. LT-4, No. 7, pp. 870-876, 1986;

B.J. Ainslie, et al. " The Fabrication and Optical Assessment of Nd 3+ in Silica-based optical Fibers", Mater. Lett., Vol. 5, No. 4, pp 143-146, 1987;

S.P. Craig-Ryan, et al. "Fabrication of Long Lengths of Low Excess Loss Erbium Doped Optical Fiber", Electron. Lett., vol. 26, No. 3, pp. 185-186, 1990;

T. Gozen, et al. "Development of High Nd +3 Content VAD

Single Mode Fibre by Molecular Stuffing Techniques", in

Proc. Optical Fibres Commun. (OFC), New Orleans, LA, 1988, Paper WQ1;

b) liquid phase techniques:

B.J. Ainslie, et al., "The Fabrication , Assessment and Optical Properties of High-concentration Nd +3 and Er+3

Doped Silica-based Fibres", Mater. Lett., Vol. 6, No. 6, pp. 139-143, 1988;

M. Nakazawa, et al. "An Ultra-efficient Erbium-doped Fiber Amplifier of 10.2 dB/mW at 0.98 um Pumping and 5.1 dB/mW at 1.48 Mm pumping", in Proc. Topical Meeting Optical Amplifiers and Applications (OSA), Monterey, CA^", Paper PdPl-1;

B.J. Ainslie, et al., "Fabrication and Optimization of the Er 3+ Distribution in Silica Based Doped Fibres", in Proc. 14th European Conf. Optical Commum. (ECOC), pp.62- 65, 1988;

J. Stone and CA. Burrus, "Nd₊₃ Doped Si0₂ Lasers in End- pumped Fiber Geometry", Appl. Phys. Lett., Vol 23, No. 7, pp. 388-389, (1973); and

c) other techniques:

J. Simpson, "Fabrication of Rare-earth Doped Fibres", in Proc. Fibre Lasers, Sources and Amplifiers Conf., 1989, pp. 2-7;

which hereby are included by reference.

5. EXAMPLES

EXAMPLE 1

"Preparation of Active Optical Fiber Waweguides"

a) Tapered Optical Waveguide

The most preferred embodiment of the invention is a tapered fiber. The outer diameter of the fiber is tapered, this will also give a tapered core radius, the core radius will vary along the length of the fiber according to the tapering of the outer diameter of the fiber.

The tapering of the fiber was done during the draw process of the fiber. When an optical fiber is drawn a diameter gauge is used to control the outer diameter of the fiber. The signal from this diameter gauge is norm¬ ally used in a feed back loop to the control system, to adjust fiber draw speed and the speed at which the pre¬ form is moved into the furnace in the drawtower in order to maintain a constant diameter of the fiber within a few micrometer.

It is no problem for those skilled in the art to use signal from this diameter gauge and the mentioned feed back loop to adjust the fiber draw speed and preform speed in such a way to obtain a fiber with any wished tapering as described in this invention.

One example of a tapered fiber was made using a preform with a uniform core having core diameter and refractive index difference constant through the length of the preform. The core was doped with Er-ions. The feed speed of the preform into the furnace of the draw tower was held constant. The draw speed of the fiber was varied in order to vary the outer diameter of the fiber. This was done by setting up to preform in the draw tower start the draw process by drawing fiber with a standard outer diameter 125 micrometer to stabilize the draw process. When the draw process was stabilized the draw speed was changed from 3 m/s to 6.5 m/s in way so that the result¬ ing fiber over a length of 300 meters had a linearly varying outer diameter changing from 125 micrometer to 85 micrometer. The straight fiber was characterized and it had a cut off at 905 nm, the absorption at 1535 nm due to the Er-ions was 0.4 dB/km and the Er-concentration was approx. 1.5 x 10 17 ions/cm3. Core diameter was 3 um.

Refractive index difference was 0.023.

In the case of tapered fibers over longer lenghts, more than 1 km, the tapering of the fiber core can also be made on the preform level. This can be done in for in¬ stance the MCVD-process, where a burner is moving along a substrate tube with a source material flow (SiCl., O^, GeCl.) in the tube. Material, mainly Si0₂, is constantly deposited on the inside wall of the mentioned tube. The deposited material is proportional to the speed of the burner.

The core profile of the fiber can now be tapered in a number of ways.

In an example the core diameter was tapered by letting the burner move in a non-constant manner. The burner was during deposition of the core layer moved with a speed changing from 100 mm/min to 50 mm/min in a way to obtain a preform with core radius changing linearly from approx¬ imately 1.5 mm to 1 mm. The core diameter can be changed by having a time varying flow of source material in the tube. Typical flows of SiCl. in this type of manufacturing are in the range from 0.1 to 1 g/min.

b) "Variable Index Difference - Straight Fiber"

The refractive index can in the same way be changed along the length of the preform by letting the amount of GeCl. vary during deposition of the core layer. Typical flows of GeCl. are in the range from 0.01 to 0.4 g/min.

c) Combinations of all these processes may be used.

EXAMPLE 2

rapered Active Optical Fiber Waveguide

An active optical fiber waveguide of-lengths between 1 and 1000 m was manufactured by tapering a preform having a longitudinally constant structure during draw.

A preform having an outer diameter of 7.5 mm and core do¬ ped with Ge in a sufficient amount to obtain a refractive index difference between core and cladding of 0.015 was provided. The core was further doped with Er in a con- centration of 2.0 10 24 ions m-3. The preform is drawn to sections of 15 - 25 m of tapered active optical fiber.

The outer diameter of the fiber was decreased during drawing of the preform from 125 um to 85 um over 20 m at a constant rate of change in fiber diameter of 2 um/m.

The tapered active optical fiber waveguide of 20 m had a 1.4 um core diameter, e.g. to be coupled to the input end of an amplifier, and a 2 um core diameter at the other end. Thus, a large extent of the mode field and a small extent of the doped region was provided at the input end. and vice versa at the output end of the active optical fiber waveguide.

The extent of mode field at the input end of the active optical fiber waveguide was close to the extent of the mode field of standard communication optical fibers.

The fiber had a gain of 30 dB at a length of 20 m for a pump power of 15 mW at 980 nm.

EXAMPLE 3

Preamplifier Using a Tapered Active Optical Waveguide

A tapered actiye optical waveguide of the type manufac¬ tured in EXAMPLE 2 wass used as a low noise active opti¬ cal fiber amplifier.

The amplifier had an input coupling efficiency of 0.85 determined by the spot size mismatch between the active fiber and a passive fiber with an RMS-spot size of 4.6 um.

The active fiber was coupled to an optical bandpass filter providing an output coupling efficiency of 0.6.

Applying a bit rate of 5 Gbit/s corresponding to an elec¬ trical bandwidth of 2.5 GHz, the receiver sensitivity was calculated for different lengths of an untapered and a tapered active fiber having a slope of 1 um/m for a counter propagating pump configuration using a pump power of 15 mW at 980 nm. Fig. 8 shows that for an untapered fiber the minimum receiver sensitivity is - 40.95dBm for a 8 m long active fiber compared to - 41.2 dBm for an tapered fiber. In order to define a measure for the tolerances within which the active fiber length can be varied for a given pump power, the 1 dB penal ry points have been marked for the untapered active fiber. We find that the interval for which the sensitivity is better than - 39.95 dBm is in¬ creased by 70 %.

It should be noted that for short lengths of both types of active fibers, there is no change in the short length limit, since the sensitivity is dominated by the ampli¬ fier shot noise. For longer lengths, however, the length limit is moved upward due to the low noise figur of the tapered amplifier. The minimum receiver sensitivity has been improved by 0.25 dB.

Further, for the tolerances within which the active fiber length can be varied, the noise figure is reduced at fib¬ er lengths above 8 m as shown in Fig. 11B at almost equal gain levels as shown in Fig. 11A.

Further, for the tolerances within which the active fiber length can be varied, the pump power at the - 40 dBm sen¬ sitivity limit is shown in Fig. 9. It is seen that the tapered active fiber have improved the length interval of the lowest pump power from about 8 to about 14 meters.

Further, for the tolerances within which the slope of the tapering can be varied, it was found that variation of the slope between 0.75 um/m and 1.25 um/m only gave rise to 10 % variation of the interval limits at a length interval defined at the - 40 dBm sensitivity limit.

Also, for the possible use in booster applications. Fig. 10 shows a gain curve for 0 dBm signal input power. It is seen that besides an increase of gain by 0.15 dB, there is a much broader tolerance within which the active fiber length can be varied. EXAMPLE 4

Experimental Comparison

In order to confirm the increased performance of the de¬ veloped fiber design, two fibers were designed and pro¬ duced.

a) Ge Co-doped Fiber, Type 1

The first fiber was produced from a standard preform with varying drawing speed in order to obtain a tapering. The fiber thus had a constant index difference and a constant confinement of 1. The index difference was 0.023. The cut-off wavelength was 910nm in the thick end and approx¬ imately 620nm in the thin end. The fiber was co-doped with Ge, thus the gain peak was at 1535nm.

At the signal output end, the mode field diameter was 5.8 um at 1550 nm and the core diameter was 3.0 um, thus the doping-radius-spotsize ratio F was about 0.52, and the outer diamter was 125 um.

At the input end, the signal mode field diamter was 6.2 um and a the doping diameter was 2.6 um, thus the F ratio was about 0.42, and the outer diamter was 110 um.

b) Al Co-doped Fiber, Type 2

The second fiber was designed and produced like the first fiber but with Al as index raising dopant instead of Ge. All other parameters were similar to the first fiber.

At the signal output end, the mode field diameter was 6.7 um at 1550 nm and the core diameter was 4.5 um, thus the doping-radius-spotsize ratio F was about 0.67, and the outer diamter was 125 um. At the input end the signal mode field diamter was 7.8 um and a the doping diameter was 3.6 um, thus the F ratio was about 0.46, and the outer diameter was 99 um.

The Al co-doping makes the ASE spectrum broader and thereby the NF measurements more precise.

The confirmation experiments were performed as a compar¬ ison between equal lengths of tapered and non tapered fibers. The non tapered fiber had an outer diameter equal to the outer diameter of the thick end of the tapered fiber.

The fibers were pumped with a Ti:Saphire laser emitting at 980nm through a wavelength selective fiber coupler as shown in the figure 18. The signal was introduced in the thin end of the fiber resulting in a co-directional pump configuration. The signal wavelength could be changed as different wavelength of a tunable DFB (Distributed Feed Back) laser with high sidemode suppression was available. The gain and noise performance was examined using a laser source of 1535nm - the gain peak of this fiber. Both the pump level and the signal level was controlled with calibrated monitor couplers spliced to the wavelength selective power combining fiber coupler. At the output of the doped fiber, a fiber pig-tail with an angled connector was spliced in order to suppress reflections and lasing in the pumped fiber. Such angled, low reflec¬ tion, optical termination was also used in the pump entrance whereas the signal input was isolated using an optical isolator.

The gain measurements were performed by intensity modula¬ tion of the signal laser source and detecting the ampli- fied signal using phase sensitive detection of the in¬ tensity modulated signal with a lock-in amplifier. The gain (in dB) is given by: = 10 log^a(Vout./'Vi.n') + Lout. ,'

where V . is the signal as measured on the lock-in amplifier after the output connector, V. is the signal measured at the point A just after the splice between the power combining coupler and the doped fiber, and L is the loss at the signal wavelength between the doped fiber and the detector in dB.

The pump power level was measured at the cut-back point A, also indicating the starting point of the fiber of length 1 under test.

The noise figure measurement was based on the ASE power density at the signal wavelength. The noise figure can be calculated from:

where P_ASE is the ASE spectral power density in W/Hz,

E . is the signal photon energy in J and G is the gain.

The NF measurements were performed using a double mono- cromator. This equipment was used to record the spectral distribution of the emitted light from the fiber end without interference from residual pump, etc.. The inser¬ tion loss of the monocromator was measured using a laser source at the center wavelength. In this way, the spec¬ tral power density could be measured, and NF could be calculated.

The gain was measured at different signal levels, differ¬ ent pump levels and the same fiberlength for the uptaper¬ ed fiber and the straight fiber.

The gain charaterization results for the uptapered fiber type 1 and the corresponding measurements for the straight iber are shown in figure 19. It can be seen that there is no significant difference between the gain curves as expected. The results of the corresponding NF measurements are shown in figure 20. It can be seen that there is a significant improvement in noise figure from the straight fiber to the uptapered fiber.

Fibers of type 2 were characterized in the same way at the same signal and pump levels. Results of the gain characterization are shown in figure 21 and the corres¬ ponding NF is shown in figure 22. It can be seen that the gain of the straight fiber in this circumstance is slightly lower than the uptapered fiber, indicating a not perfectly optimized fiberlength situation. Most important is the result of the NF, which shows significant improvement compared to the straight fiber.

The results shown in figs. 19-22 can be compared to the theoretically calculated performance. The result of the calculation of the gain is seen in figure 23. It is worth mentioning that the calculated gain compares well with the measured gain. The calculated noise figures are shown in figure 24. Note the expected improvement of 0.5 - 1.0 dB of the noise figure just as measured.

Claims

P A T E N T C L A I M S :

1. A method of amplifying a light signal propagating in an optical waveguide, comprising the steps of:

a) interrupting the waveguide to establish spaced apart a first and a second waveguide end section;

b) passing the light signal from the first waveguide end section to an active optical waveguide extending longitudinally from a first end to a second end com¬ prising:

ii) a cladding (11,21) surrounding the guiding region and having a refractive index less than the re¬ fractive index of the guiding region,

whereby the mode fields of the propagating light signal and the pump light are determined by the extent of the guiding region, the refractive index difference between the guiding region and the cladding, or both,

c) generating signal light amplification in the active optical waveguide for at least a range of wavelengths including the bandwidths of the light signal by pumping light energy into the active optical wave¬ guide; and

d) passing the amplified light signal from the active optical waveguide to the second waveguide end section; CHARACTERIZED in that

the active optical waveguide has a longitudinally vari¬ able structure in which at least one of the parameters:

2. A method according to claim 1, CHARACTERIZED in that the F-ratios vary in the range from 0.1 to 0.8, prefer¬ ably from 0.25 to 0.8.

3. A method according to claims 1 - 2, CHARACTERIZED in that the light signal is introduced into the active opti¬ cal waveguide co-propagating with the pump light.

4. A method according to claims 1 - 2, CHARACTERIZED in that the light signal is introduced into the active opti¬ cal waveguide counter-propagating with the pumping light.

5. A method according to claims 1 - 4, CHARACETERIZED in that the extent of guiding region and the extent of active dopant is increasing monotonically from the end where the pump light is introduced to the other end and that the refractive index difference is constant.

6. A method according to claims 1 - 4, CHARACTERIZED in that the extent of guiding region and the extent of active dopant are constant and the refractive index difference is increasing monotonically from the end where the pump light is introduced to the other end.

7. A method according to claims 1 - 4, CHARACTERIZED in that the extent of guiding region and the refractive index difference are constant and that the extent of active dopant is increasing monotonically from the end where the pump light is introduced to the other end.

8. A method according to claims 1 - 7, CHARACTERIZED in that the active dopant is a fluorescent additive, parti¬ cularly a rare earth element selected from the group consisting of Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu, preferably Er, Nd, Pr, Ho, Yb, and Tm, most pre- ferably Er.

9. A method according to any of claims 1 - 8, CHARACTER¬ IZED in that the active optical waveguide includes the active dopant in a concentration from 5-10 to 1*10 _3 ions m , preferably for distributed amplifiers from

1-10 20 to 10-1020 ions m-3, and preferablby for short o oc __o amplifiers from 1-10 to 2-10 ions

10. An active optical waveguide suitable for amplifying a light signal having a predetermined wavelength, com¬ prising:

ii) a cladding (11,21) surrounding the guiding region and determining the extent thereof, and having a refractive index less than the refractive index of the guiding region, whereby the mode fields of the propagating light signal and the pump light are determined by the extent of the guiding region, the refractive index difference between the guiding region and the cladding, or both,

CHARACTERIZED in that

11. An active optical waveguide according to claim 10, CHARACTERIZED in that the F-ratios vary in the range from 0.1 to 0.8, preferably from 0.25 to 0.8.

12. An active optical waveguide according to claims 10 - 11, CHARACETERIZED in that the extent of guiding region and the extent of active dopant is increasing monoton¬ ically from the first end to the second end of the active optical waveguide and that the refractive index differ¬ ence is constant.

13. An active optical waveguide according to claims 10 - 11, CHARACTERIZED in that the extent of guiding region and the extent of active dopant are constant and the refractive index difference is increasing monotonically from the first end to the second end of the active opti- cal waveguide.

14. An active optical waveguide according to claims 10 - 11, CHARACTERIZED in that the extent of guiding region and the refractive index difference are constant and that the extent of active dopant is increasing monotonically from the first end to the second end of the active opti¬ cal waveguide.

15. An active optical waveguide according to claims 10 -

14, CHARACTERIZED in that the active dopant is a fluore¬ scent additive, particularly a rare earth element selec¬ ted from the group consisting of Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu, preferably Er, Nd, Pr, Ho, Yb, and Tm, most preferably Er.

16. An active optical waveguide according to claims 10 -

15, CHARACTERIZED in that the active optical waveguide includes the active dopant in a concentration from 5-10 19 c _ to 1*10 ions m^~ , preferably for distributed amplifiers from 1-10 20 to 10^•1020 ions m-3, and preferablby for short amplifiers from 1-10 to 2-10 ions m

17. An active optical waveguide according to claims 10 - 15, CHARACTERIZED in that the waveguide is an optical fiber waveguide.

18. An active optical waveguide according to claims 10 - 17, CHARACTERIZED in that the optical fiber waveguide is a tapered optical waveguide.

19. A method of manufacturing an active optical fiber waveguide suitable for amplifying a light signal having a predetermined wavelength, CHARACTERIZED by: a) providing a preform having a longitudinally constant structure with parameter values selected within the fol¬ lowing ranges:

ii) a constant transverse extent of dopant in the full extent of the core;

iii) a constant refractive index difference between the

_3 core aanncd cladding in the range from 5-10 to

20. A method of manufacturing an active optical fiber waveguide suitable for amplifying a light signal having a predetermined wavelength, CHARACTERIZED by:

i) a variable transverse extent of the core providing a final core diameter from 0.5 to 20 um of the active optical fiber waveguide, said extent of the final core varying a factor of 1.1 to 4 from the one end of the final active optical fiber waveguide to the other;

iii) a constant refractive index difference between the

_3 core and cladding in the range from 5-10 to

50-10^"3;

21. A method of manufacturing an active optical fiber waveguide suitable for amplifying a light signal having a predetermined wavelength, CHARACTERIZED by:

i) a constant transverse extent of the core in the ran¬ ge providing a final core radius from 0.5 to 20 um of the active optical fiber waveguide;

ii) a constant transverse extent of dopant in the range from 0.05 to 1.00 times the core diameter; iii) a variable refractive index difference between the core and cladding varying a factor of 1.1 to 4 from the one end of the preform to the other in the range from 5*10^"3 to 50-10^"3;

22. A method of manufacturing an active optical fiber waveguide suitable for amplifying a light signal having a predetermined wavelength, CHARACTERIZED by:

a) providing a preform having a longitudinally variable structure with parameter values selected within the fol¬ lowing ranges: _ -

i) a constant transverse extent of the core providing a final core diameter from 0.5 to 20 um of the active optical fiber waveguide;

iii) a constant refractive index difference between the core and cladding in the range from 5-10 -3 to 50*10^"3; b) drawing the preform into the active optical fiber waveguide with a length of 1 - 100 km; and

23. A method of manufacturing an active optical fiber waveguide suitable for amplifying a light signal having a predetermined wavelength, CHARACTERIZED by:

i) a variable transverse extent of the core providing a final core diameter from 0.5 to 20 um of the active optical fiber waveguide, sai:⁷ extent of the final core varying a factor of 1.1 „o 4 from the one end of the final active optical fiber waveguide to the other;

iii) a variable refractive index difference between the core and cladding varying a factor of 1.1 to 4 from the one end of the preform to the other in the range from 5-10^"3 to 50-lθ^'3;

b) drawing the preform into the active optical fiber waveguide with a length of 1 - 100 km; and c) selecting the values of the above mentioned parameters to ensure that the values of the ratio F of the extent of active dopant to the extent of the mode field of the signal light fall within the range from 0.05 to 1.00.

24. An optical amplifier, comprising:

f) coupling means for coupling the amplified light signals out of the active optical waveguide;

CHARACTERIZED in that

is varied from a first end of the active optical wavegui¬ de to a second end in such a way that the ratio F of the extent of active dopant to the extent of the mode field of the signal light is increased monotonically from the first end to the second end, the values of said ratio F being selected within the range from 0.05 to 1.00. 25. An optical amplifier according to claim 24, CHARAC¬ TERIZED in that the F-ratios vary in the range from 0.1 to 0.8, preferably from 0.

25 to 0.8.

26. An optical amplifier according to claims 24 - 25,

CHARACETERIZED in that the extent of guiding region and the extent of active dopant is increasing monotonically from the first end to the second end of the active optical waveguide and that the refractive index differ- ence is constant.

27. An optical amplifier according to claims 24 - 25, CHARACTERIZED in that the extent of guiding region and the extent of active dopant are constant and the re- fractive index difference is increasing monotonically from the first end to the second end of the active optical waveguide.

28. An optical amplifier according to claims 24 - 25, CHARACTERIZED in that the extent of guiding region and the refractive index difference are constant and that the extent of active dopant is increasing monotonically from the first end to the second end of the active optical waveguide.

29. An optical amplifier according to claims 24 - 28, CHARACTERIZED in that the active dopant is a fluorescent additive, particularly a rare earth element selected from the group consisting of Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu, preferably Er, Nd, Pr, Ho, Yb, and Tm, most preferably Er.

30. An optical amplifier according to claims 24 - 29,

CHARACTERIZED in that the active optical waveguide in- eludes the active dopant in a concentration from 5^•10 19 f. _o to 1*10 ions m , preferably for distributed amplifiers from 1*10 20 to 10-1020 ions m-3, and preferably for short amplifiers from 1-10 to 2-10

31. An optical amplifier according to claims 24 - 30, CHARACTERIZED in that the waveguide is an optical fiber waveguide.

32. An optical amplifier according to claims 24 - 31, CHARACTERIZED in that the optical fiber waveguide is a tapered optical waveguide.

33. An optical amplifier according to claims 24 - 32 in which the source of light signals comprises an optical fiber transmitting the signals, CHARACTERIZED in that the transverse mode field of the active optical waveguide is matched to the transverse mode field of the fiber.

34. An optical amplifier according to claims 24 - 33, CHARACTERIZED in that the transverse mode field of the coupling means is matched to the transverse mode field of the active optical waveguide.

35. An optical amplifier according to claims 24 - 34, CHARACTERIZED in that the coupling means comprises wave¬ length selective means.

36. An optical amplifier according to claims 24 - 35, CHARACTERIZED in that it further comprises a second active optical waveguide, the second end of which is connected to the second end of the first active optical waveguide, and the first end of the second active optical waveguide is connected to the coupling means for coupling the amplified light signals out of the second active optical waveguide, and

that the pumping light source is coupled to the first end of the active waveguide.

37. A fiber optical transmission system, comprising:

CHARACTERIZED in that one or more of the amplifiers are amplifiers according to claims 24 - 36.