CN1305242C - WDM system channel power dynamic equalization method equalizing optical amplifier thereof - Google Patents

Abstract

The present invention relates to a wavelength division multiplex (WDM) system channel power dynamically equalizing method and an equalizing optical amplifier thereof. The method is characterized in that the method integrates gain spectrum flatness, a gain amplitude regulating mechanism and optical amplification into a whole by utilizing a linear relationship among a channel power variation, pumping power and an optical attenuation dependent variable, and multichannel output power is maintained to be equalized and constant while the dynamic gain spectrum flatness is realized; the method which can be independently designed according to system anticipate parameters is simultaneously suitable for the dynamic power equalization of a point-to-point multiway system and the dynamic power equalization of an optical network. The equalizing optical amplifier adopts an isolated cascade amplifying optical path to suppress amplified spontaneous radiation; a loss compensating filter is introduced to flatten the gain spectrum; an adjustable optical attenuator is introduced to regulate the gain amplitude. The equalizing optical amplifier controlled by a single chip computer responds to a variation of channel input power and/or a variation of the number of channels to dynamically and linearly regulate and control the pumping power and the optical attenuation variable, and thereby, the dynamic equalization of multiway channel output optical power is realized.

Description

The method of wavelength-division multiplex system channel power dynamic equalization and balanced image intensifer thereof

Technical field

The method of wave division multiplexing WDM system channel power dynamic balance and balanced image intensifer thereof belong to high-speed wideband optical fiber communication and optical amplifier technology field, particularly the erbium-doped fiber amplifier EDFA of the method for channel power dynamic equalization and power equalization thereof in the WDM Networks of Fiber Communications.

Background technology

The appearance of EDFA (1986) once caused the once great leap of Fibre Optical Communication Technology development, and the WDM transmission technology that adopts EDFA to amplify becomes the dominant direction of current optical fiber communication engineering system development.Along with the rapid growth of multiple communication services such as data, voice, image, multimedia in recent years, optical communication system is constantly towards the direction develop rapidly of two-forty, big capacity, long distance.In addition, the carrying out of Added Business such as the use of broadband user's access device and interactive multimedia transmission is that the wdm optical communication network of target becomes in recent years " focus " of research and development both at home and abroad to realize the high-speed high capacity broad band Integrated service.

The dynamic operation of wdm optical communication network requires image intensifer that the transmission channel of a plurality of wavelength is carried out the equilibrium amplification, and so-called " dynamic equalization " specifically comprises here:

(1) to determining a plurality of constant power optical channel (λ in the transmission bandwidth _i) when amplifying simultaneously, should be able to provide essentially identical gain of light G (λ for each channel _i); When analog optical signal was amplified, differential gain dG (λ)/d λ should be as far as possible little, and promptly image intensifer should have smooth gain spectral.

(2) when channel number changes, should keep constant flat gain spectrum (being called the gain strangulation); When channel power change or channel power and channel number change simultaneously, should adjust the amplitude of flat gain spectrum automatically and keep constant channel output power spectrum (being called power limiting).

Owing to the intrinsic unevenness of Er-doped fiber EDF gain spectral, especially cause the competition of interchannel based on the spectral property of HOMOGENEOUS BROADENING, the gain difference that channel obtained of different wave length, and this gain inequality also can amplify with cascade to be accumulated, and will produce the transmission characteristic of wdm system to have a strong impact on:

(1) in the point-to-point wdm system of long distance, the peak value that the deflection long wavelength appears in the gain spectral that the EDFA cascade amplifies also degenerates signal to noise ratio, even can cause some channel gain to increase severely and other channel is suppressed.

(2) along with the development of technology such as Optical Add Drop Multiplexer OADM, the cross interconnected OXC of light, wdm system to the point of new generation that has the light exchange capacity-multiple spot Networks of Fiber Communications development, requires to possess functions such as network dynamic restructuring, fault self-recovery by simple Point-to-Point system.In fiber optic network, the variation of routed path can make each channel power that enters node EDFA change; In addition, the channel number by EDFA also may with on the reconstruct of network or the channel/change down, cause the dynamic distortion of EDFA gain spectral and output power spectrum on the node.

(3) in containing the mixing wdm system of unlike signal form, power lower (＜-(＞3dBm) the vision signal conllinear transmission of 10dBm) baseband digital signal and higher-wattage, not only can cause the huge differential of interchannel gain and power, in case and video signal interruption, EDFA promptly jumps to the small-signal state from dark saturation condition, causes the transmission of digital channel to suffer unacceptable power fluctuation.

This shows, for guaranteeing the dynamic operation of wdm optical communication network, should be under the situation that channel number and channel power all may change, EDFA should keep the flatness of gain spectral, simultaneously can keep each to be exaggerated the constant of channel power output again, promptly realize the dynamic equalization of multichannel through-put power, this is the key technology difficult point that the wdm optical communication network moves towards the engineering practicality.

The inventor had once described the method that realizes EDFA gain spectrum flattening and strangulation simultaneously in the patent (patent No. 97 2 248965.5) of " realizing the method and the balanced image intensifer of wdm system dynamics of channels gain balance ": employing is inserted the loss balancing filter and is made the planarization of static gain spectrum in EDFA; Based on the physical mechanism of gain spectral, by two kinds of strangulations that method realizes gaining of electric feedback regulation pump power or light feedback regulation ideler frequency laser power by the decision of averaged particles counter-rotating degree.

Yet the method for above-mentioned channel gain dynamic equalization all is that the variation feedback regulation number average particle counter-rotating degree of the multichannel gross power of foundation input EDFA makes it to remain unchanged, thereby keeps the constant of flat gain spectrum.This only changes and channel power when keeping constant when channel number, and the flat gain spectrum of strangulation could guarantee to remain the power output equilibrium of channel.And when channel power itself changes, though the flat gain spectrum of strangulation can guarantee each channel and obtain same gain that the channel power output will change.Therefore, not only the variation that simple gain strangulation technology can not compensate for channel power, and in the system that the EDFA cascade is amplified, also constantly accumulation and influence the normal working point of subsequent stages of the variation of this channel power finally can be caused the severe exacerbation of system transmissions quality.

So far, the technology of the channel power equilibrium of existing some realization fiber optic transmission systems.For example:

(1) in optical cable television CATV system, EDFA only amplifies individual channel usually, can utilize the automatic gain adjusting function under the saturated magnifying state of its degree of depth, and the optical channel that input power is changed is realized the constant of power output by the automatic adjusting that gains.

Yet for the EDFA that uses in wdm system, the number average particle counter-rotating degree that the input channel variable power causes changes the also distortion thereupon of shape that will make gain spectral, will cause the unbalance of each channel gain and power, so this method and inapplicable.

(2) in wdm system, add lump dynamic power equalizer every certain transmission range place that several cascades EDFA amplifies, the channel power spectrum of distortion is concentrated shaping, the power of each transmission channel is equated.At present, this dynamic power equalizer mainly contains two kinds:

A) demultiplexing/multiplexer DMUX/MUX+ adjustable optical attenuator VOA or EDFA array: the flashlight of the multi-wavelength multiplex in the Transmission Fibers resolves into a plurality of channels of single wavelength by DMUX, VOA or EDFA array regulate the optical attenuation of each channel or luminous power that the gain of light makes each channel equates, gets back to Transmission Fibers after a plurality of single wavelength channel through adjusting closes ripple by MUX again.

B) many audio frequencies cascaded optical fiber acousto-optic filter AOTF: on optical fiber, load sound wave and can constitute band resistance optical filter, but adopt the cascaded optical fiber acousto-optic filter broadening filtered band of a plurality of different frequency of sound wave excitations, regulating acoustic power can form and the antisymmetric compound acousto-optic filtering spectrum of transmission channel power spectrum, make the higher fading channel of power bigger, and the lower fading channel of power is less or be not attenuated, can be with the channel power spectrum leveling of distortion.

These the two kinds methods that adopt lump dynamic power equalizer from the angle of transmission system overall operation are except complex structure, equipment cost costliness, and subject matter is:

The first, from operation principle, the lump dynamic power equalizer that adopts VOA array or AOTF is the equilibrium that realizes the channel power spectrum by the optical attenuation of regulating each channel respectively, the former is a cost to sacrifice channel light power, thereby must compensate with additional image intensifer.

The second, from design and compatibility, the transmission performance and the practical operation situation of system that the structural design of lump dynamic power equalizer and parameter dispose and portfolio effect places one's entire reliance upon, for different transmission systems, must carry out special design respectively, thereby device compatible relatively poor to system.

Three, from using and managing, the complexity the when system of employing lump dynamic power equalizer must carry out a large amount of field adjustables and operation is controlled and is safeguarded, has further increased the complexity and the operation cost of this technology.

Summary of the invention

The objective of the invention is to weak point, propose a kind of new method that solves wdm system channel power dynamic equalization problem at prior art.Its basic ideas are: since will keep the dynamically smooth of EDFA gain spectral, just must keep its number average particle counter-rotating degree constant; And to realize the strangulation of channel power output, just must realize the dynamic regulation of flat gain spectral amplitude.Therefore, the present invention proposes to set up the smooth and gain range regulatory mechanism of dynamic gain spectrum in image intensifer inside, constitutes the dynamic equalization image intensifer, realizes the strangulation of dynamic power when realizing the planarization of dynamic gain spectrum.

For illustrating the present invention composes smooth and gain range regulate and control method about the EDFA dynamic gain operation principle, need under number average particle degree of counter-rotating steady state, carry out theory analysis to the decay dependent variable of channel input power change amount, VOA and the relation between the pump power dependent variable.

Consider that two sections Er-doped fiber EDF1 and the EDF2 cascade of interpolation adjustable optical attenuator VOA shown in Figure 1 and gain flattening filter GFF amplify the light path schematic diagram.The amplified spont-aneous emission ASE of stable state Giles transmission equation and population equation transmit, ignore to(for) multichannel, non-single direction are:

$\frac{{\mathrm{dP}}_k\left(z\right)}{\mathrm{dz}}=u_k{P}_k\left(z\right)\{({\mathrm{\&alpha;}}_k+g_k^{*})\frac{N_2\left(z\right)}{N_T}-({\mathrm{\&alpha;}}_k+l_k)\}+u_{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">k</figure-callout>}}2{\mathrm{hv}}_k\mathrm{\&Delta;}{v}_k{g}_k^{*}\frac{N_2\left(z\right)}{N_T}---\left(1\right)$

$\frac{N_2\left(z\right)}{N_T}=-\frac{1}{\mathrm{\&zeta;}}\underset{k}{\mathrm{\&Sigma;}}{u}_k\frac{{\mathrm{dP}}_k\left(z\right)}{{\mathrm{hv}}_k\mathrm{dz}}---\left(2\right)$

Wherein, h is a Planck's constant, v _kBe frequency of light wave, P _k(z) representing frequency is v _kLuminous power, N ₂(z) be last energy level erbium ion concentration, N _TBe erbium ion total concentration, u _kGet+1 and-1 represent respectively light edge+Z and-Z direction transmission, l _kBe the background loss of EDF, α _k, g _k ^*Be respectively absorption coefficient, gain coefficient and the saturation coefficient of EDF with ζ.

Under saturated amplification condition of work, establish the loss of VOA and the equal and Wavelength-independent of supplementary load loss of other elements, derive the total net gain G of channel of whole EDFA by (1), (2) two formulas _k ^TotFor:

$G_k^{\mathrm{tot}}=G_{1k}\&times;\frac{1}{A}\&times;G_{2k}=\frac{1}{A}\frac{1}{A_{1k}^{\mathrm{add}}{A}_{2k}^{\mathrm{add}}}[({\mathrm{\&alpha;}}_k+g_k^{*})\frac{{\stackrel{\&OverBar;}{N}}_2}{N_T}-({\mathrm{\&alpha;}}_k+l_k)L]=\frac{1}{A}{G}_{\mathrm{sk}}---\left(3\right)$

L=L in the formula ₁+ L ₂Be EDF total length, L ₁And L ₂Be respectively the EDF length of prime and back level, k=1,2 ... M represents channel ordinal number, G _IkAnd A _Ik ^AddBeing respectively with VOA is gain and the supplementary load loss of the forward and backward two-stage (i=1,2) of boundary to k channel, and A is the attenuation of VOA, $\stackrel{\&OverBar;}{{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">N</figure-callout>}}_2}=\frac{1}{L}{\&Integral;}_0^L{N}_2\left(z\right)\mathrm{dz}$ Be the integration of erbium ion among the EDF along fiber lengths L, N _TBe erbium ion total concentration, α _k, g _k ^*And l _kBe respectively absorption coefficient, gain coefficient and the background loss of EDF, G _Sk=G _1k* G _2kBe the overall gain of k channel of whole EDFA, under the flat gain spectral condition, each channel gain is roughly the same, so the approximate G that gets _Sk=G _sBe the channel overall gain of whole EDFA except that the VOA attenuation.

(3) formula shows that total net gain size of EDFA is by forward and backward two-stage gain G _1kAnd G _2kAnd VOA attenuation A decision, but the shape of gain spectral a number average particle counter-rotating degree N by two-stage EDF ₂/ N _TDecision.Thereby as long as keep N ₂/ N _TConstant, can keep the flatness of gain spectral.Obtain the relation of number average particle counter-rotating degree and channel power and the pump power of whole EDFA along z vector product branch by (2) formula:

$-\mathrm{L\&zeta;}\frac{{\stackrel{\&OverBar;}{N}}_2}{N_T}={\frac{1}{{\mathrm{hv}}_s}\mathrm{MP}}_s^{\mathrm{in}}(\frac{G_s}{A}-1+(1-\frac{1}{A}){\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{Av}})-\frac{1}{{\mathrm{hv}}_{p1}}{\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1-\frac{1}{{\mathrm{hv}}_{p2}}{\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2---\left(4\right)$

${\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{Av}}=\frac{1}{M}\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{G}_{1k}$

H is a Planck's constant in the formula, v _P1Be prime pumping light frequency, v _P2Be back level level pump light frequency, v _sBe the flashlight frequency, ζ is the saturation coefficient of EDF, P _s ^InBe average channel input power, P _P1 ^InAnd P _P2 ^InBe respectively the input pump power of prime and back level, G ₁ ^AvAverage gain for prime.By (4) formula as seen, the size of the position at VOA place and attenuation A thereof all can influence the population inversion degree of whole EDFA.

1, the relation of the decay dependent variable of channel input power change amount and VOA

When the channel input power changes, regulate the attenuation of VOA, to keep the constant of channel power output; Keep channel overall gain G by regulating pump power _sConstant, then obtain VOA decay change amount Δ A and channel input power change amount Δ P by (3) formula _s ^InBetween relation:

$\frac{P_s^{\mathrm{in}}}{A}=\frac{P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{A-\mathrm{\&Delta;A}}\&DoubleRightArrow;\mathrm{\&Delta;A}=A\frac{{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{P_s^{\mathrm{in}}}---\left(5\right)$

Promptly when adjusting the VOA attenuation to keep P _s ^InWhen/A is constant, the Δ A and the Δ P that represent with dB _s ^InValue equates that this just obtains the quantitative control relation of VOA.

2, the relation of channel input power change amount and prime pump power dependent variable

The initial condition of supposing EDFA is the also maximum situation of attenuation of input power maximum (being channel input power and number of channel maximum), corresponding VOA.As channel input power P _s ^InChange and number of channel M when constant: ${\mathrm{MP}}_s^{\mathrm{in}}\&RightArrow;M(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}}),$ Determine the attenuation A-Δ A of VOA to regulate the prime pump power simultaneously by (5) formula $P_p^{}$ $1_{\mathrm{in}}^{}\&RightArrow;{\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1-{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}1}^{\mathrm{in}},$ Make prime population inversion degree (thereby gain G ₁ ^Av) constant.So, utilize (4), (5) formula to obtain prime pump power dependent variable Δ P _P1 ^InWith channel input power change amount Δ P _s ^InRelation:

$\frac{1}{{\mathrm{<figure-callout\; id="1"\; label="hv\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}1}^{\mathrm{in}}=\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">M</figure-callout>}\frac{1}{{\mathrm{hv}}_s}{\mathrm{\&Delta;P}}_s^{\mathrm{in}}\&times;{\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{Av}}---\left(6\right)$

As seen Δ P _P1 ^InWith Δ P _s ^InRelation only with the prime gain G ₁ ^AvRelevant, and with overall gain G _sIrrelevant.

3. input channel is counted the relation of change amount and back level pump power dependent variable

As channel power P _s ^InWhen changing simultaneously with channel number M: ${\mathrm{MP}}_s^{\mathrm{in}}\&RightArrow;(M+\mathrm{\&Delta;M})(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}}),$ On aforementioned basis, regulate back level pump power $P_p^{}$ $2_{\mathrm{in}}^{}\&RightArrow;{\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2-{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}2}^{\mathrm{in}}.$ At this moment, because of prime pumping power P _P1 ^In-Δ P _P1 ^InNo longer regulate, then preceding stage gain will become G ₁ ^Av+ Δ G ₁ ^AvSo, utilize (4), (5) and (6) formula to obtain back level pump power dependent variable Δ P _P2 ^InRelation with number of channel change amount Δ M:

$\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="hv\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}2}^{\mathrm{in}}=\mathrm{\&Delta;<figure-callout\; id="1"\; label="M"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">M</figure-callout>}\frac{1}{{\mathrm{hv}}_s}{P}_s^{\mathrm{in}}\frac{G_s}{A}---\left(7\right)$

As seen Δ P _P2 ^InOnly become with Δ M, and with channel input power change amount Δ P _s ^InAnd the decay dependent variable Δ A of VOA is irrelevant.

(5), (6) and (7) formula shows: channel input power change amount Δ P _s ^InAnd between the decay dependent variable Δ A of VOA, channel input power change amount Δ P _s ^InWith prime pump power dependent variable Δ P _P1 ^InBetween, input channel counts change amount Δ M and back level pump power dependent variable Δ P _P2 ^InBetween all keep good linear relationship.

Method of the present invention is characterised in that:

It is that a kind of smooth and gain range regulation mechanism of dynamic gain spectrum of setting up simultaneously in image intensifer inside constitutes the dynamic equalization image intensifer, so that when realization dynamic gain spectrum is smooth, realize the method for power limiting, promptly under Single-chip Controlling, in the EDFA of at least two sections Er-doped fiber EDF cascade that its light path part is separated with an adjustable optical attenuator VOA, adopt the method for light-electricity regulation and control to regulate the flatness and the gain range of dynamic gain spectrum, promptly with fiber coupler from amplifier in extension flashlight, after receiving, photodetector is transformed into the signal of telecommunication, regulate attenuation and the front and back level pump power of VOA with it, so that when guaranteeing gain spectrum flattening, realize the strangulation of channel power by regulating gain range, make multiplex (MUX) in input power or/and channel number keeps Output optical power impartial and constant when changing; The method of described dynamic equalization is to adjust according to the different situations of channel power and channel number respectively, and they contain following steps respectively successively:

(1). when channel power changes, with the change amount Δ P of photodetector monitoring individual channel input power _s ^In, calculate the respective attenuation dependent variable Δ A of the VOA that should be provided with according to following formula, in order to regulate the gain of EDFA:

$\mathrm{\&Delta;A}=A\frac{{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{P_s^{\mathrm{in}}},$

Wherein, A is the attenuation of VOA, P _s ^InMean Input Power for channel; Simultaneously calculate corresponding prime pump power dependent variable Δ P by following formula _P1 ^In, and the prime Pump Drive current placed corresponding value, to keep number average particle counter-rotating degree constant, promptly the flatness of gain spectral is constant, thereby realizes the dynamic strangulation that channel power is composed;

$\frac{1}{{\mathrm{<figure-callout\; id="1"\; label="hv\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}1}^{\mathrm{in}}=\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">M</figure-callout>}\frac{1}{{\mathrm{hv}}_s}{\mathrm{\&Delta;P}}_s^{\mathrm{in}}\&times;{\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}},$

Wherein, M is a total number of channels, and h is a Planck's constant, v _P1Be prime pumping light frequency, v _sBe flashlight frequency, G ₁ ^AVBe preceding stage gain, ${\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}=\frac{1}{M}\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{G}_{1k},$ G _1kFor with VOA being of the gain of the prime of boundary to k channel;

(2). when channel number changes, monitor corresponding total channel input power change amount with photodetector ${\mathrm{\&Delta;P}}_s^{\mathrm{tot}}=P_s^{\mathrm{in}}\&times;\mathrm{\&Delta;M},$ And calculate channel number change amount Δ M in view of the above, obtain the dependent variable Δ P of back level pump power by following formula _P2 ^In, and a level Pump Drive current in back is arranged at corresponding value, to keep number average particle counter-rotating degree constant, promptly realize the strangulation that gains, thereby guarantee that each channel power output is constant;

$\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="hv\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}2}^{\mathrm{in}}=\mathrm{\&Delta;<figure-callout\; id="1"\; label="M"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">M</figure-callout>}\frac{1}{{\mathrm{hv}}_s}{P}_s^{\mathrm{in}}\frac{G_s}{A},$

Wherein, G _sBe the overall gain of channel, A is the attenuation of adjustable optical attenuator VOA;

(3). when channel power and channel number change simultaneously,, utilize the change amount Δ P of the individual channel input optical power of monitoring with the change amount of photodetector monitor AMP input optical power _s ^In, VOA is set in corresponding decay dependent variable Δ A with said method, calculate corresponding prime pump power dependent variable Δ P by said method again _P1 ^In, and the prime Pump Drive current is arranged at corresponding value; Utilize the change amount of the total channel input power of monitoring ${\mathrm{\&Delta;P}}_s^{\mathrm{Tot}}={\mathrm{\&Delta;P}}_s^{\mathrm{in}}\&times;M+P_s^{\mathrm{in}}\&times;\mathrm{\&Delta;M}$ Change amount Δ P with described individual channel input power _s ^In, the change amount Δ M that calculates the input channel number obtains the dependent variable Δ P of back level pump power _P2 ^InAnd a level Pump Drive current is arranged at analog value after inciting somebody to action, thereby the collaborative adjusting by the front and back stages pump power keeps number average particle counter-rotating degree constant, promptly keeps under the state of gain spectrum flattening, realizes the dynamic strangulation of channel output power spectrum by the adjusting of gain spectral amplitude.

For the wave division multiplexing WDM system, the changes delta P of single communication channel input power _s ^InObtain by following method monitoring: the service channel that is provided with the 1510nm wavelength usually follows multiplexing communication channel to propagate in optical fiber, these two kinds of channels should be followed identical power decay rule in transmission course, measure the power change amount Δ P of this service channel in EDFA with photodetector ₁₅₁₀After, just can release the change amount Δ P of single communication channel input power by following formula _s ^In:

$\frac{{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{P_s^{\mathrm{in}}}=\frac{{\mathrm{\&Delta;P}}_{1510}}{{\mathrm{<figure-callout\; id="1510"\; label="P"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_{1510}}.$

The balanced image intensifer that the present invention proposes is characterised in that:

The EDFA of at least two sections Er-doped fiber EDF cascade that the light path part of whole amplifier is separated with adjustable optical attenuator VOA, be positioned at the one section Er-doped fiber that links to each other with pumping source along optical path direction before the VOA and constitute prime EDFA, and be positioned at the EDF formation back level EDFA more than a section or a section that links to each other with pumping source after the VOA.

As shown in Figure 2, it contains wavelength division multiplexing optical fiber wave multiplexer MUX21, Er-doped fiber EDF21, optical isolator ISO22, adjustable optical attenuator VOA, wave multiplexer MUX22, Er-doped fiber EDF22, optical isolator ISO23, fiber grating filter GFF, wave multiplexer MUX23, Er-doped fiber EDF23, wave multiplexer MUX24 and the Er-doped fiber EDF24 that is provided with in turn along the input signal light path; The pumping input of wave multiplexer MUX21-MUX24 respectively with pumping source Pump21, Pump22, Pump23, when Pump24 links to each other, the free end 212 of wave multiplexer MUX21 is connected optical isolator ISO21 and ISO24 respectively with the free end 204 of Er-doped fiber EDF24, and is used separately as the input and the output of amplifier signal light; For the variation of monitoring light input signal luminous power to realize dynamic regulation, input 201 at optical isolator ISO21 connects optical fiber channel-splitting filter Tap22 and Tap21, wherein 5% of Tap22 partial wave end 203 connects photodetector PIN22, the variable power of monitoring 1525-1625nm band communication channel; 5% the partial wave end 202 of Tap21 connects photodetector PIN21, the variable power of monitoring 1510nm wavelength service channel.

Described balanced image intensifer is to set up the fiber amplifier that dynamic gain is composed smooth and gain range regulation mechanism simultaneously in inside.

Described balanced image intensifer adopts two sections of the power division pumpings of interpolation optical isolator ISO or the cascade of multistage EDF to amplify light path.Described ISO places prime (i.e. the 1st section EDF) afterwards, and its effect is that the reverse ASE that stops back level EDF enters prime EDF, makes cascade amplifier also have high-gain, high power and low-noise characteristic in the channel power strangulation that realizes than great dynamic range.

It is to amplify in the cascade of two sections of the power division pumpings of described interpolation ISO or multistage EDF to insert the gain flattening filter GFF with special type loss spectra in the light path that the dynamic gain of described balanced image intensifer is composed smooth mechanism.Described GFF is its loss spectra becomes inverting to distribute with the gain spectral evolving trend a broadband optical filter, is positioned at ISO appropriate location afterwards, adjusts and its filter is put down by the static gain spectrum of GFF loss spectra pair amplifier.

The gain range regulation mechanism of described balanced image intensifer is to amplify in the cascade of two sections of the power division pumpings of described interpolation ISO and GFF or multistage EDF to insert adjustable optical attenuator VOA in the light path.Described VOA is that its light decrement can be by regulating the adjustable optical attenuator that its driving voltage changes, its attenuation adjustable extent and amplifier total channel input power excursion decibel (dB) value are quite, be positioned at after the ISO of isolating amplifier prime and back level, it is a key function element of realizing the amplifier channel power dynamic balance.

The outstanding advantage of this method is: the first, compose dynamic gain smooth and gain range regulation mechanism and light amplification process combine together, it is minimum that Power penalty is dropped to, need not to dispose in addition image intensifer and carry out power back-off, structure is greatly simplified, equipment cost reduces; The second, can independently carry out the parameter comprehensive Design according to the transmission characteristic of system's expection, not rely on the apparent situation of system parameter, thereby different systems be had compatible preferably; Three, integral structure and the operating characteristic thereof because of the dynamic equalization image intensifer is not subjected to the system parameter dynamic effects, thereby its portfolio effect is not subjected to the influence of system running state or its position, has absolute equilibrium or with meeting balanced characteristics, both be applicable to point-to-point wdm system, also be suitable for the WDM optical-fiber network.

Description of drawings

Fig. 1 is an EDFA dynamics of channels power equalization principle schematic of the present invention.

Fig. 2 is the light channel structure schematic diagram of the balanced EDFA embodiment of dynamic power of the present invention.

Fig. 3 is the test result of embodiment of the invention prime Pump Drive current and channel input power relation.

Fig. 4 is the test result that level Pump Drive current and equivalent input power concerns after the embodiment of the invention.

(a) relation curve of level pump power and input channel number after during VOA attenuation A=15.6

(b) relation curve of level pump power and input channel number after during VOA attenuation A=1

Fig. 5 is the test result of the gain spectral dynamic equalization of the balanced EDFA of the embodiment of the invention,

1: gain G=36.2dB, decay A=0dB,---Pin=32 μ w,----Pin=16.21 μ w ... Pin=8.12 μ w

2: gain G=33.3dB, decay A=3dB,---Pin=64 μ w,----Pin=16.1 μ w

3: gain G=30.2dB, decay A=6dB,---Pin=125 μ w,----Pin=6.5 μ w

4: gain G=27.2dB, decay A=9dB,---Pin=248 μ w,----Pin=125.2 μ w

5: gain G=24.3dB, decay A=12dB,---Pin=501 μ w,----Pin=125.6 μ w ... Pin=75.1 μ w

Fig. 6 is a hardware principle block diagram of the present invention

Fig. 7 is the program flow diagram of single-chip microcomputer among the present invention

Embodiment

By light channel structure shown in Figure 2, wavelength division multiplexing optical fiber wave multiplexer MUX21, Er-doped fiber EDF21, optical isolator ISO22, adjustable optical attenuator VOA, wave multiplexer MUX22, Er-doped fiber EDF22, optical isolator ISO23, fiber grating filter GFF, wave multiplexer MUX23, Er-doped fiber EDF23, wave multiplexer MUX24, and Er-doped fiber EDF24 is linked in sequence the pumping input 211 of wave multiplexer MUX21, the pumping input 221 of MUX22, the pumping end 231 of MUX23 and the pumping input 241 of MUX41 respectively with pumping source Pump21, Pump22, Pump23 links to each other with Pump24.In order to prevent the influence of light reflection, the free end 212 of wave multiplexer MUX21 is connected optical isolator ISO21 and ISO24 respectively with the free end 204 of doped fiber EDF24, and is used separately as the input and the output of amplifier signal light; For the variation of monitoring light input signal luminous power to realize dynamic regulation, input 201 at optical isolator ISO21 connects optical fiber channel-splitting filter Tap22 and Tap21, wherein 5% of Tap22 partial wave end 203 connects photodetector PIN22, the variable power of monitoring 1525-1625nm band communication channel; 5% the partial wave end 202 of Tap21 connects photodetector PIN21, the variable power of monitoring 1510nm wavelength service channel.Whole amplifier is that the boundary will be divided into two-stage with adjustable optical attenuator VOA, and prime is made of the Er-doped fiber EDF21 of Pump21 pumping, and the back level comprises Pump22, Pump23 and Pump24 Er-doped fiber EDF22, EDF23 and the EDF24 of pumping respectively.

The course of work of this EDFA is as follows: pump light injects four sections doped fibers from four optical fiber wave multiplexers respectively, forms area of popular inversion in Er-doped fiber; Flashlight enters first section Er-doped fiber EDF21 amplification through optical fiber channel-splitting filter Tap21, optical isolator ISO21 and wave multiplexer MUX21, amplify through optical isolator ISO22, adjustable optical attenuator VOA, wave multiplexer MUX22, doped fiber EDF22 secondary then, flashlight after secondary amplifies is by optical isolator ISO23 and through fiber grating filter GFF filtering, filtered flashlight priority is through wave multiplexer MUX23 and MUX24 enters the 3rd section Er-doped fiber EDF23 and the 4th section Er-doped fiber EDF24 amplifies once more, after optical isolator ISO24 output.The filtering spectrum of fiber grating filter GFF and Amplifier Gain spectrum antisymmetry wherein is used for the gain spectral of smooth EDFA; Photodetector PIN21 and PIN22 convert the variable power of the 1510nm wavelength service channel surveyed and the variable power of 1525-1625nm wave band input signal light to the signal of telecommunication respectively, are used for adjustable optical attenuator VOA and forward and backward two-stage pump power.

The erbium that uses in the enforcement/aluminium codoped optical fiber numerical aperture 0.21, cut-off wavelength 0.9mm is respectively 3.29dB/m and 4.57dB/m to the absorption of pump light and flashlight.Four sections long respectively 12m of Er-doped fiber EDF1, EDF22, EDF23 and EDF24,7m, 3.6m and 7m adopt 980nm laser diode pumping source, and the power of Pump21, Pump22, Pump23 and Pump24 is respectively 160mW, 80mW, 80mW and 230mW.

Because the actual pump power that injects Er-doped fiber can't be measured, measurable physical quantity is a Pump Drive current, can utilize the linear relationship between Pump Drive current and the pump power, by measuring the variation delta I of Pump Drive current _Pi ^InThe variation delta P that replaces pump power _Pi ^InIn addition, pump light has nothing to do by the general size with pump power of loss that coupler injects Er-doped fiber, thereby can think the variation delta I of Pump Drive current _Pi ^InVariation delta P with the channel input power _s ^InReach number of channel change amount Δ M and also should satisfy the linear relationship that is similar to (5) and (6) two formulas, just differ a constant coefficient, the linear relationship that this coefficient can be measured between Pump Drive current and the channel input power is by experiment determined.For guaranteeing the precision of control, can measure many group channel input powers and Pump Drive current value in the experiment and carry out linear fit, under different channel input powers and channel number condition, control pump power then with the linear relationship that match obtains.

When measuring gain spectral, with an ASE wide range source analog input light signal, but this can only simulate total input optical power, simultaneously the change of analog channel number and channel power.For this reason, suppose and know the channel input power,, change the variation that the channel input power is come the analog channel number then with this drive current of setting the attenuation of VOA and setting prime pumping Pump21.

At first, by the control strategy that (6) formula provides, the power of prime pumping Pump21 is just thought just to regulate when the channel input power changes.In experiment, to import ASE wide range source power and drop to 32 μ W by 500 μ W, setting the maximum channel number is 32, then the respective channels input power (18dBm) changes to 1 μ W (30dBm) by 15.6 μ W, the gain spectral of monitoring prime output, regulate prime pumping Pump21 power gain spectral remained unchanged, the driving current value that records prime pumping Pump21 and the relation of channel input power as shown in Figure 3, as seen the data of surveying present extraordinary linearity:

I _p1(mA)＝18.97×(P _sin/channel)+31.35 (8)

By the control strategy that (7) formula provides, back level pump power is only regulated with the change of the number of channel, and is irrelevant with the channel input power, so can measure under the channel input power arbitrarily.Experiment measuring channel input power maximum (18dBm) and minimum value (30dBm) carry out under (just the VOA attenuation is maximum 12dB and minimum value 0dB) two kinds of situations, and get the control relation of both mean value as reality.When adopting ASE wide range source to come the analog input channel, the number of channel can be calculated as follows:

$M=\frac{P_s^{\mathrm{in}}}{P_{s,\mathrm{chennal}}^{\mathrm{in}}}=\frac{1}{{\left(P_{s,\mathrm{channel}}^{\mathrm{in}}\right)}_{\min }}\frac{P_s^{\mathrm{in}}}{A}---\left(9\right)$

P wherein _s ^InAnd P _{S, channel} ^InBe respectively total power input and individual channel power.For simplicity, definition P _s ^In/ A is equivalent input power, and measures the relation of back level Pump Drive current and equivalent input power.

Be given in experiment records under the constant situation of EDFA number average particle counter-rotating degree the back grade of Pump Drive current and the relation of equivalent input power of keeping among Fig. 4.Because the restriction of pump power, the excursion (excursion of the number of channel) of equivalence input power has exceeded the control range of any one pumping of back level, need regulate and to meet the demands two pumpings, so curve has comprised the control relation of Pump23 and two pumpings of Pump24 among the figure.When the channel input power was maximum [being that the VOA attenuation is maximum A=15.6 (11.93dB)], the back level Pump Drive current that experiment records and the relation of equivalent input power were shown in Fig. 4 (a), and the control relation that obtains is:

$I_{p3}\left(\mathrm{mA}\right)=\left\{\begin{array}{cc}24.12\&times;(P_{\sin }/A)-13.57& P_{\sin }/A\&le;6\\ {\left(I_{p3}\right)}_{\max }& P_{\sin }/A\&GreaterEqual;6\end{array}\right.$

$I_{p4}\left(\mathrm{mA}\right)=\left\{\begin{array}{cc}0& P_{\sin }/A\&le;6\\ 15.39\&times;(P_{\sin }/A)-83.66& P_{\sin }/A\&GreaterEqual;6\end{array}\right.---\left(10\right)$

When the channel input power was minimum value [the VOA attenuation is minimum value A=1 (0dB)], the back level Pump Drive current that experiment records and the relation of equivalent input power were shown in Fig. 4 (b), and the control relation that obtains is:

$I_{p3}\left(\mathrm{mA}\right)=\left\{\begin{array}{cc}22.85\&times;(P_{\sin }/A)-21.24& P_{\sin }/A\&le;6\\ {\left(I_{p3}\right)}_{\max }& P_{\sin }/A\&GreaterEqual;6\end{array}\right.$

$I_{p4}\left(\mathrm{mA}\right)=\left\{\begin{array}{cc}0& P_{\sin }/A\&le;6\\ 15.14\&times;(P_{\sin }/A)-86.81& P_{\sin }/A\&GreaterEqual;6\end{array}\right.---\left(11\right)$

Get the control relation of the mean value of (10) and (11) formula as back level pumping:

$I_{p3}\left(\mathrm{mA}\right)=\left\{\begin{array}{cc}23.49\&times;(P_{\sin }/A)-17.41& P_{\sin }/A\&le;6\\ {\left(I_{p3}\right)}_{\max }& P_{\sin }/A\&GreaterEqual;6\end{array}\right.$

$I_{p4}\left(\mathrm{mA}\right)=\left\{\begin{array}{cc}0& P_{\sin }/A\&le;6\\ 15.27\&times;(P_{\sin }/A)-85.24& P_{\sin }/A\&GreaterEqual;6\end{array}\right.---\left(12\right)$

At last, utilize (8) formula and (12) formula under the condition of different channels input power and channel number, measure the gain spectral of EDFA.In experiment, the input power of winning the confidence is respectively-30dBm ,-27dBm ,-24dBm ,-21dBm and-18dBm, earlier by the attenuation of this condition enactment VOA; Control pumping Pump21 by (8) formula again, change input power, calculate corresponding equivalent total power input P _s ^In/ A sets pumping Pump23 and pumping Pump24 according to (12) formula, measures many group gain spectral, as shown in Figure 5.As can be seen, utilize control method of the present invention, under the situation of whole amplifier gain-variable, realized the regulation and control of flat gain spectrum.With the 1550nm wavelength is example, and when the channel input power changed, control precision reached Δ G＜0.2dB; When signal number changed, control precision reached Δ G＜0.3dB.

Hardware principle block diagram and Single Chip Microcomputer (SCM) program flow chart are seen Fig. 6, Fig. 7.

This shows that the present invention has following advantage:

(1) described wdm system channel power dynamic equalization new method integrates dynamic gain spectrum smooth and gain range regulation mechanism and light amplification process, can reduce to need not to dispose in addition image intensifer and carry out power back-off because of introducing the Power penalty that gain of light regulation mechanism is brought.

(2) the described wdm system channel power dynamic equalization new method comprehensive Design that can independently carry out parameter according to the transmission parameter of system expection realizes, and does not rely on the apparent situation of system parameter, and different systems is had compatibility preferably.

(3) described wdm system channel power dynamic equalization new method is not subjected to the system parameter dynamic effects, its portfolio effect is not subjected to the influence of system running state or its position, thereby has absolute equilibrium or with meeting balanced characteristics, both be applicable to point-to-point multiloop loop system, also be suitable for the dynamic equalization of optical network system channel power.

(4) described wdm system channel power dynamic equalization new method is not subjected to the restriction of service band, both has been applicable to the system that uses C-wave band EDFA, is applicable to the system that uses L-wave band EDFA and other doped optical fibre amplifiers yet.

The derivation of specification (5) (6) (7) formula

Consider erbium-doped fiber amplifier basic structure shown in the Figure of description 1.

At first calculate the gain of whole amplifier.With VOA is separation, and amplifier is divided into two parts, is first before the VOA, is second portion afterwards.(2) formula in the specification is obtained two-part net gain along EDF1 and EDF2 integration respectively is:

$G_{1i}=\frac{1}{A_{1i}^{\mathrm{add}}}\&times;\exp \left[(({\mathrm{\&alpha;}}_{\mathrm{si}}+g_{\mathrm{si}}^{*}){\&Integral;}_0^{L1}\frac{N_2}{N_T}\mathrm{dz}-({\mathrm{\&alpha;}}_{\mathrm{si}}+l_{\mathrm{si}}))L_1\right]$ (attached 1)

$G_{2i}=\frac{1}{A_{2i}^{\mathrm{add}}}\&times;\exp [(({\mathrm{\&alpha;}}_{\mathrm{si}}+g_{\mathrm{si}}^{*}){\&Integral;}_{L1}^{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">L</figure-callout>}1+L2}\frac{N_2}{N_T}\mathrm{dz}-({\mathrm{\&alpha;}}_{\mathrm{si}}+l_{\mathrm{si}}))({\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+L_2)]$ (attached 2)

α wherein _Si, g _Si ^*And l _SiBe respectively absorption coefficient, gain coefficient and the background loss of EDF under i channel wavelength, N ₂(z) be last energy level erbium ion concentration, N _TBe erbium ion total concentration, A _1i ^Add, A _2i ^AddBe the supplementary load loss of first and second portion, comprise the welding loss, the loss of GFF loss and other device.

Between two sections, because the attenuation of VOA, make that the input power of the power output of amplifier first and second portion is different, both satisfy relations of plane down:

${\mathrm{<figure-callout\; id="2"\; label="P\; si"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_{\mathrm{si}}^{}=\frac{1}{\mathrm{<figure-callout\; id="1"\; label="A\; P\; si"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">A</figure-callout>}}{P}_{\mathrm{si}}^{}{}_{1\mathrm{stageout}}^{}=\frac{G_{1i}}{A}{P}_{\mathrm{si}}^{\mathrm{in}}$ (attached 3)

Wherein, A is the attenuation multiple of VOA under linear coordinate, is an amount with Wavelength-independent.

By (attached 1), (attached 2) and (attached 3) formula obtains whole amplifier and to the gain of signal is:

$G_i^{\mathrm{Tot}}=G_{1i}\&times;\frac{1}{A}\&times;G_{2i}$

$=\frac{1}{A}\&times;\frac{1}{A_{1i}^{\mathrm{add}}{A}_{2i}^{\mathrm{add}}}\exp \left[(({\mathrm{\&alpha;}}_{\mathrm{si}}+g_{\mathrm{si}}^{*})\frac{\stackrel{\&OverBar;}{{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">N</figure-callout>}}_2}}{N_T}-({\mathrm{\&alpha;}}_{\mathrm{si}}+l_{\mathrm{si}}))L\right],i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,M$

$=\frac{1}{A}\&times;G_{\mathrm{si}}$ (attached 4)

Wherein: G _Si=G _1i* G _2iBe whole amplifier overall gain in corresponding wavelength except that VOA, usually, in the service band of EDFA, can think that the overall gain of each wavelength is identical, i.e. G _Si=G _s

Calculate the number average particle counter-rotating degree of amplifier below again.With in (1) formula in the specification

Along the whole amplifier length integration and (attached 3) formula of application, ignore the spontaneous radiation item The signal power of residue pump power and input obtains the average counter-rotating degree of whole amplifier and the pass of signal power and pump power and is:

$-\mathrm{L\&zeta;}\frac{\stackrel{\&OverBar;}{{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">N</figure-callout>}}_2}}{N_T}=$

$\underset{i}{\mathrm{\&Sigma;}}\frac{1}{{\mathrm{hv}}_{\mathrm{si}}}\left\{P_{\mathrm{si}}^{\mathrm{in}}\right[G_{1i}-1]+{\mathrm{<figure-callout\; id="2"\; label="P\; si"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_{\mathrm{si}}^{}[G_{2i}-1\left]\right\}-\frac{1}{{\mathrm{hv}}_{p1}}[{\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1-{\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1]-\frac{1}{{\mathrm{hv}}_{p2}}[{\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2-{\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2]$

$=\frac{1}{{\mathrm{hv}}_s}{\mathrm{MP}}_s^{\mathrm{in}}[\frac{G_s}{A}-1]+\frac{1}{{\mathrm{hv}}_s}{\mathrm{MP}}_s^{\mathrm{in}}(1-\frac{1}{A})\&times;{\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{Av}}-\frac{1}{{\mathrm{hv}}_{p1}}{\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1-\frac{1}{{\mathrm{hv}}_{p2}}{\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2$ (attached 5)

Wherein: G _sBe the overall gain of amplifier, and ${\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}{G}_{1i}$ Average gain for amplifier first.The input power of also having supposed each channel of amplifier simultaneously equates, thereby total input power is written as MP _s ^In, M is a channel number, P _s ^InMean Input Power for channel.Such hypothesis is the situation that meets the practical application of wdm system.

The control of attached 1. variable attenuators

In the method for the invention, when the channel input power changes, the attenuation of VOA will change thereupon, to keep the channel power output constant, can be obtained by (attached 4) formula:

$P_s^{\mathrm{out}}=P_s^{\mathrm{in}}\&times;\left[\frac{G_s}{A}\right]=[P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}}]\&times;\left[\frac{G_s}{A-\mathrm{\&Delta;A}}\right]$ (attached 6)

Here, can think when input power and VOA attenuation change, keep G by regulating pumping _sConstant, then can obtain:

$\mathrm{\&Delta;A}=A\frac{{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{P_s^{\mathrm{in}}}$ (attached 7)

The quantitative control relation of (attached 7) formula VOA just illustrates (5) formula in the book.

The control of attached .2 pump power

For the convenience of deriving, the variation of input power was divided into for two steps:

1. the first step: ${\mathrm{MP}}_s^{\mathrm{in}}\&RightArrow;M(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}}),$ Have only the channel input power to change, and channel number is constant.

2. second go on foot: $M(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}})\&RightArrow;(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}})(M-\mathrm{\&Delta;M}),$ On the channel input power basis of the first step, channel number changes.

For the convenience of deriving, when the initial condition of supposing amplifier was the input power maximum, just the channel input power and the number of channel were maximum, and corresponding VOA attenuation also is a maximum.

In the first step, have only the channel input power to change, at this moment determine the attenuation of VOA by (attached 7) formula, the power of regulating pumping 1 simultaneously is P _P1 ^In-Δ P _P1 ^In, with the gain G of hold amplifier first ₁ ^AVConstant.So, the input power that can obtain second portion EDF should be

With the power before the adjusting

Equate that this moment, the pump power of second portion did not change, the counter-rotating degree of second portion EDF will remain unchanged so, thus the average counter-rotating degree of whole amplifier With total net gain G _sAlso with constant.With MP in (attached 5) formula _s ^InReplace with M (P _s ^In-Δ P _s ^In), A replaces with A-Δ A, P _P1 ^InReplace with P _P1 ^In-Δ P _P1 ^In, and application (attached 7) obtains:

$-\mathrm{L\&zeta;}\frac{\stackrel{\&OverBar;}{{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">N</figure-callout>}}_2}}{N_T}=\frac{1}{{\mathrm{hv}}_s}{\mathrm{MP}}_s^{\mathrm{in}}\frac{G_s}{A}+\frac{1}{{\mathrm{hv}}_s}M(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}}){\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}$

$-\frac{1}{{\mathrm{hv}}_{p1}}({\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1-{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}1}^{\mathrm{in}})-\frac{1}{{\mathrm{hv}}_{p2}}{\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2$ (attached 8)

(attached 8) formula and (attached 5) formula are just subtracted each other and can be obtained:

$\frac{1}{{\mathrm{<figure-callout\; id="1"\; label="hv\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}1}^{\mathrm{in}}=\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">M</figure-callout>}\frac{1}{{\mathrm{hv}}_s}{\mathrm{\&Delta;P}}_s^{\mathrm{in}}\&times;{\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}$ (attached 9)

(attached 9) formula is the variable quantity of first's pump power and the relation between the channel input power variable quantity, that is to say (6) formula in the specification.

In second step, the channel input power keeps the value after the first step changes, and channel number changes, and is P with the power adjustments of second portion pumping _P2 ^In-Δ P _P2 ^In, at this moment, because the input power of first section EDF changes, and first section pump power keeps the performance number after the first step changes, and changes again, and first section EDF gain will become G so ₁+ Δ G ₁With M (P in (attached 8) formula _s ^In-Δ P _s ^In) replace with (M-Δ M) (P _s ^In-Δ P _s ^In), P _P2 ^InReplace with P _P2 ^In-Δ P _P2 ^InObtain:

$-\mathrm{L\&zeta;}\frac{\stackrel{\&OverBar;}{{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">N</figure-callout>}}_2}}{N_T}=\frac{1}{{\mathrm{hv}}_s}(M-\mathrm{\&Delta;M})P_s^{\mathrm{in}}\frac{G_s}{A}$

$+\frac{1}{{\mathrm{hv}}_s}(M-\mathrm{\&Delta;M})(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}})({\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}+{\mathrm{\&Delta;<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}})$

$-\frac{1}{{\mathrm{hv}}_{p1}}({\mathrm{<figure-callout\; id="1"\; label="P\; p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^1-{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}1}^{\mathrm{in}})-\frac{1}{{\mathrm{hv}}_{p2}}({\mathrm{<figure-callout\; id="2"\; label="P\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">P</figure-callout>}}_p^2-{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}2}^{\mathrm{in}})$ (attached 10)

(attached 10) formula and (attached 8) formula are subtracted each other the relation that obtains between second portion pump power variable quantity and the channel number variable quantity:

$\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="hv\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}2}^{\mathrm{in}}=\frac{1}{{\mathrm{hv}}_s}{\mathrm{\&Delta;MP}}_s^{\mathrm{in}}\frac{G_s}{A}$

$-\frac{1}{{\mathrm{hv}}_s}(P_s^{\mathrm{in}}-{\mathrm{\&Delta;P}}_s^{\mathrm{in}})[(M-\mathrm{\&Delta;M})({\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}+{\mathrm{\&Delta;<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}})-{\mathrm{<figure-callout\; id="1"\; label="MG"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">MG</figure-callout>}}_1^{\mathrm{AV}}]$ (attached 11)

In in (attached 11) last, M (P _s ^In-Δ P _s ^In) G ₁ ^AVIt is the power output of first.At the output of single hop amplifier, when not considering input signal power, and the residue pump power is negligible the time, and the power of output will be a constant value, and this is the operate in saturation characteristic of EDF just.When the EDF of first is operated in saturation condition, M (P _s ^In-Δ P _s ^In) G ₁ ^AVTo be a steady state value, and in this step the input power P of channel _s ^In-Δ P _s ^InRemain unchanged, so, $(M-\mathrm{\&Delta;M})({\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}}+{\mathrm{\&Delta;<figure-callout\; id="1"\; label="G"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">G</figure-callout>}}_1^{\mathrm{AV}})-{\mathrm{<figure-callout\; id="1"\; label="MG"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">MG</figure-callout>}}_1^{\mathrm{AV}}=0.$ Then (attached 11) formula can be reduced to:

$\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="hv\; p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">hv</figure-callout>}}_p}{\mathrm{\&Delta;P}}_{\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="C0215349100202.png"\; state="\{\{state\}\}">p</figure-callout>}2}^{\mathrm{in}}=\mathrm{\&Delta;<figure-callout\; id="1"\; label="M"\; filenames="C0215349100182.png,C0215349100202.png"\; state="\{\{state\}\}">M</figure-callout>}\frac{1}{{\mathrm{hv}}_s}{P}_s^{\mathrm{in}}\frac{G_s}{A}$ (attached 12)

(attached 12) formula is exactly the relation of the channel number of the pump power of second portion and input, that is to say (7) formula in the specification.

Claims (4)

1, the method of wavelength-division multiplex system channel power dynamic equalization, contain the step that realizes that simultaneously erbium-doped fiber amplifier EDFA gain spectrum flattening and gain range are regulated, it is characterized in that: it is that a kind of smooth and gain range regulation mechanism of dynamic gain spectrum of setting up simultaneously in image intensifer inside constitutes the dynamic equalization image intensifer, so that when realization dynamic gain spectrum is smooth, realize the method for power limiting, promptly under Single-chip Controlling, in the EDFA of at least two sections Er-doped fiber EDF cascade that its light path part is separated by an adjustable optical attenuator VOA, adopt the method for light-electricity regulation and control to regulate the flatness and the gain range of dynamic gain spectrum, promptly with fiber coupler from amplifier in extension flashlight, after receiving, photodetector is transformed into the signal of telecommunication, regulate attenuation and the front and back level pump power of VOA with it, so that when guaranteeing gain spectrum flattening, realize the strangulation of channel power by regulating gain range, make multiplex (MUX) in input power or/and channel number keeps Output optical power impartial and constant when changing; The method of described dynamic equalization is to adjust according to the different situations of channel power and channel number respectively, and they contain following steps respectively successively:

(1). when channel power changes, with the change amount Δ P of photodetector monitoring individual channel input power _s ^In, calculate the respective attenuation dependent variable Δ A of the VOA that should be provided with according to following formula, in order to regulate the gain of EDFA:

$\mathrm{\&Delta;A}=A\frac{{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{P_s^{\mathrm{in}}},$

Wherein, A is the attenuation of VOA, P _s ^InMean Input Power for channel; Simultaneously calculate corresponding prime pump power dependent variable Δ P by following formula _P1 ^In, and the prime Pump Drive current placed corresponding value, to keep number average particle counter-rotating degree constant, promptly the flatness of gain spectral is constant, thereby realizes the dynamic strangulation that channel power is composed;

$\frac{1}{{\mathrm{hv}}_{p1}}{\mathrm{\&Delta;P}}_{p1}^{\mathrm{in}}=M\frac{1}{{\mathrm{hv}}_s}{\mathrm{\&Delta;P}}_s^{\mathrm{in}}\&times;G_1^{\mathrm{AV}},$

Wherein, M is a total number of channels, and h is a Planck's constant, v _P1Be prime pumping light frequency, v _sBe flashlight frequency, G ₁ ^AVBe preceding stage gain, $G_1^{\mathrm{AV}}=\frac{1}{M}\underset{k=1}{\overset{M}{\mathrm{\&Sigma;}}}{G}_{1k},$ G _1kFor with VOA being of the gain of the prime of boundary to k channel;

(2). when channel number changes, monitor corresponding total channel input power change amount with photodetector ${\mathrm{\&Delta;P}}_s^{\mathrm{tot}}=P_s^{\mathrm{in}}\&times;\mathrm{\&Delta;M},$ And calculate channel number change amount Δ M in view of the above, obtain the dependent variable Δ P of back level pump power by following formula _P2 ^In, and a level Pump Drive current in back is arranged at corresponding value, to keep number average particle counter-rotating degree constant, promptly realize the strangulation that gains, thereby guarantee that each channel power output is constant;

$\frac{1}{{\mathrm{hv}}_{p2}}{\mathrm{\&Delta;P}}_{p2}^{\mathrm{in}}=\mathrm{\&Delta;M}\frac{1}{{\mathrm{hv}}_s}{P}_s^{\mathrm{in}}\frac{G_s}{A},$

Wherein, G _sBe the overall gain of channel, A is the attenuation of adjustable optical attenuator VOA;

(3). when channel power and channel number change simultaneously,, utilize the change amount Δ P of the individual channel input optical power of monitoring with the change amount of photodetector monitor AMP input optical power _s ^In, VOA is set in corresponding decay dependent variable Δ A with said method, calculate corresponding prime pump power dependent variable Δ P by said method again _P1 ^In, and the prime Pump Drive current is arranged at corresponding value; Utilize the change amount of the total channel input power of monitoring ${\mathrm{\&Delta;P}}_s^{\mathrm{Tot}}={\mathrm{\&Delta;P}}_s^{\mathrm{in}}\&times;M+P_s^{\mathrm{in}}\&times;\mathrm{\&Delta;M}$ Change amount Δ P with described individual channel input power _s ^In, calculate the change amount Δ M of input channel number, obtain the dependent variable Δ P of back level pump power again by the formula in the above-mentioned steps (2) _P2 ^InAnd a level Pump Drive current is arranged at analog value after inciting somebody to action, thereby the collaborative adjusting by the front and back stages pump power keeps number average particle counter-rotating degree constant, promptly keeps under the state of gain spectrum flattening, realizes the dynamic strangulation of channel output power spectrum by the adjusting of gain spectral amplitude.

2, the method for wavelength-division multiplex system channel power dynamic equalization according to claim 1 is characterized in that: for the wave division multiplexing WDM system, and the changes delta P of single communication channel input power _s ^InObtain by following method monitoring: the service channel that is provided with the 1510nm wavelength follows multiplexing communication channel to propagate in optical fiber, these two kinds of channels should be followed identical power decay rule in transmission course, measure the power change amount Δ P of this service channel in EDFA with photodetector ₁₅₁₀After, just can release the change amount Δ P of single communication channel input power by following formula _s ^In:

$\frac{{\mathrm{\&Delta;P}}_s^{\mathrm{in}}}{P_s^{\mathrm{in}}}=\frac{{\mathrm{\&Delta;P}}_{1510}}{P_{1510}}.$

3, the method for wavelength-division multiplex system channel power dynamic equalization according to claim 1 and the balanced image intensifer that designs, it is characterized in that: the EDFA of at least two sections Er-doped fiber EDF cascade that the light path part of whole amplifier is separated with adjustable optical attenuator VOA, be positioned at the one section Er-doped fiber that links to each other with pumping source along optical path direction before the VOA and constitute prime EDFA, and be positioned at the EDF formation back level EDFA more than a section or a section that links to each other with pumping source after the VOA.

4, balanced image intensifer according to claim 3, it is characterized in that: it contains along the input signal light path wavelength division multiplexing optical fiber wave multiplexer MUX21 is set in turn, Er-doped fiber EDF21, optical isolator ISO22, adjustable optical attenuator VOA, wave multiplexer MUX22, Er-doped fiber EDF22, optical isolator ISO23, fiber grating filter GFF, wave multiplexer MUX23, Er-doped fiber EDF23, wave multiplexer MUX24, and Er-doped fiber EDF24, the pumping input of wave multiplexer MUX21～MUX24 respectively with pumping source Pump21, Pump22, Pump23, when Pump24 links to each other, the free end (204) of the free end of wave multiplexer MUX21 (212) and Er-doped fiber EDF24 is connected optical isolator ISO21 and ISO24 respectively, and is used separately as the input and the output of amplifier signal light; For the variation of monitoring input signal light power to realize dynamic regulation, input (201) at optical isolator ISO21 connects optical fiber channel-splitting filter Tap22 and Tap21, wherein, 5% the partial wave end (203) of Tap22 connects photodetector PIN22, the variable power of monitoring 1525-1625nm band communication channel; 5% the partial wave end (202) of Tap21 connects photodetector PIN21, the variable power of monitoring 1510nm wavelength service channel.

Images