JP2004151227A - Gain equalizer, optical amplifier, and optical transmission system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a gain equalizing technique with which the wavelength characteristics of the light are efficiently compensated with high accuracy by combining a plurality of periodic filters. <P>SOLUTION: The gain equalizer 1a has loss wavelength characteristics exhibiting single wave increasing or decreasing with respect to the wavelength region for compensation and is equipped with a first compensating part 11 to compensate the linear gain component of the wavelength characteristics in the wavelength region for compensation and a second compensating part 12 to compensate the wavelength characteristics in the wavelength region for compensation not compensated in the first compensating part 11 with an optical filter 12A consisting of a plurality of the combined periodic filters cascade connected to the first compensating part 11. <P>COPYRIGHT: (C)2004,JPO

Description

【００５２】
ここで、８０波のＷＤＭ信号光に対して各々の光周波数の信号光が存在するか否かを表す次の（２）式のような重み行列Ｗを導入し、最小二乗線形フィッティング法を適用する。
【００５３】
【数２】

【００５４】
ただし、上記（２）式のｗ_ｉについては、信号光が存在する時にｗ_ｉ＝１、存在しない時にｗ_ｉ＝０とする。
【００５５】
ここで、上記の（１）式に関して次の（３）式に示す行列Ａ，ｂ，ｘを定義する。
【００５６】
【数３】

【００５７】
そして、上記の（１）式〜（３）式の関係を用いて重みつき最小二乗方程式を表すと、次の（４）式のようになる。
【００５８】
【数４】

【００５９】
この（４）式をｘについて解くと次の（５）式となる。
【００６０】
【数５】

【００６１】
従って、（５）式の行列式を計算することにより、ＣおよびＤの値を求めることができる。
【００６２】
上記のような最小二乗線形フィッティング法による（５）式の関係を基に、光モニタ部１５でモニタされる各光周波数ν_ｉに対応した光パワーＰ_ｉまたは光信号対雑音比ＯＳＮＲ_ｉの値を用いて、光増幅器の一次チルト成分が計算により求められる。ここでは、具体的な計算過程の一例として、図９に示したスペクトルについて、図１０に示すように各光周波数ν_ｉに対応した光パワーＰ_ｉをそれぞれ読み取り、各々の値を用いて光増幅器の一次チルト成分を計算した結果を示しておく。
【００６３】
まず、図１０から読み取った各値を前述の（１）式に代入すると、次の（６）式のようになる。
【００６４】
【数６】

【００６５】
また、図９に示したようにここではすべての光周波数の信号光が存在するため、前述の（２）式におけるｗ_１〜ｗ_８０の値はすべて１となる。これらの関係より、前述の（５）式を演算してｘを求めると、次の（７）式のようになる。
【００６６】
【数７】

【００６７】
従って、図９に示したモニタ結果に対して、光増幅器の利得の一次チルト成分は、図１０中の実線で示すように、一次傾斜値Ｄ＝０．８９３０７５（ｄＢ／Ｈｚ）を持つ直線で近似される。
【００６８】
次に、上記のような一次チルト成分が第１補償部１１’で補償されるようにフィルタ特性の最適化が図られる。
【００６９】
具体的に、ここでは第１補償部１１’に用いられる線形光フィルタ１１Ａ’として、例えば、利得等化を行う周波数帯域の２倍以上のＦＳＲをもつ長周期フィルタを使用する場合を考える。この場合、長周期フィルタの最適化パラメータはフィルタ特性の振幅および位相となる。このため、長周期フィルタの振幅および位相と一次チルト成分の傾斜値Ｄとの関係式を求めて、フィルタ特性の最適化を行うことが必要となる。このような線形光フィルタ１１Ａ’の最適化は、例えば、以下に示すような最小二乗線形フィッティング法を適用した計算手順に従って行うことが可能である。
【００７０】
線形光フィルタ１１Ａ’として用いられる長周期フィルタの理論式は、純粋な周期フィルタの理論式で近似することができるとして、次の（８）式のような関係で表される。
【００７１】
【数８】

【００７２】
ただし、ｓはフィルタの振幅（真値）、ΔνはフィルタのＦＳＲ（ＴＨｚ）、νは光周波数、ν_ｂはフィルタのボトム周波数（ＴＨｚ）である。なお、上記のボトム周波数は、フィルタの減衰量が０ｄＢになるときの周波数を示す値である。
【００７３】
上記の長周期フィルタについても、前述した場合と同様にして、一次式：ｙ＝Ｃ’＋Ｄ’×ｘという関係式を当てはめると、Ｆ（ν_ｉ）＝Ｃ’＋Ｄ’×ν_ｉとなる。この関係式について、信号光の存在する８０波に対して最小二乗線形フィッティング法を適用し、次の（９）式に示す行列式を利用する。
【００７４】
【数９】

【００７５】
なお、第１補償部１１’のフィルタ特性の最適化に関しては、すべての光周波数に対してＦ（ν_ｉ）は存在するとして設計を行うことができるため、前述の（２）式に示したような重み行列Ｗを導入する必要はない。
【００７６】
ここで、上記の（９）式について次の（１０）式に示す行列Ａ，ｂ’，ｘ’を定義する。
【００７７】
【数１０】

【００７８】
上記の（９）式および（１０）式の関係を用いて最小二乗方程式を表すと、ｂ’＝Ａｘ’となり、この最小二乗方程式をｘ’について解くと次の（１１）式となる。
【００７９】
【数１１】

【００８０】
この（１１）式の行列式を計算することで、Ｃ’およびＤ’の値を求めることができる。
【００８１】
また、周期フィルタの性質から明らかなように、フィルタの位相を制御することによって、所要の振幅のフィルタから一次傾斜の絶対値が等しい２つの線形光フィルタを実現することができる。このため、前述の（５）式に従って計算された光増幅器の一次チルト成分の傾きの絶対値｜Ｄ｜と、上記の（１１）式に従って計算される周期フィルタの傾きの絶対値｜Ｄ’｜とが等しくなるように、フィルタの振幅ｓの最適化が図られる。実際には、長周期フィルタの振幅ｓと一次傾斜の絶対値Ｄ’との関係式を求め、その関係式に従って｜Ｄ｜＝｜Ｄ’｜が成り立つように線形光フィルタ１１Ａ’の振幅ｓを最適化すればよい。
【００８２】
長周期フィルタの振幅と一次傾斜の絶対値との関係式は、一般的に複雑なものとなる。このため、（５）式に従って計算された一次傾斜の絶対値｜Ｄ｜からフィルタの振幅ｓを最適化する際には、制御部１６において例えばニュートン法などによる探索を行うことが好適である。また、フィルタの位相については、フィルタの平均挿入損失を最小にするため、（５）式に従って計算された一次傾斜値Ｄが正の場合には、光周波数ν_１とフィルタのボトム周波数ν_ｂとが等しくなるようにし（ν_１＝ν_ｂ）、一方、一次傾斜値Ｄが負の場合には、光周波数ν_８０とフィルタのボトム周波数ν_ｂとが等しくなり（ν_８０＝ν_ｂ）となるように制御するのが望ましい。
【００８３】
ここでは、前述の具体例で計算した一次傾斜値Ｄ＝０．８９３０７５に対応させて、線形光フィルタ１１Ａ’の振幅ｓの最適化を図った計算過程を一例として示しておく。
【００８４】
例えば、線形光フィルタ１１Ａ’として用いる長周期フィルタのＦＳＲΔνを１２ＴＨｚ、ボトム周波数ν_ｂを１９１．９５ＴＨｚとすると、上記の（１１）式は、次の（１２）式のようになる。
【００８５】
【数１２】

【００８６】
上記の（１２）式より、｜Ｄ｜＝｜Ｄ’｜とするためには、次の（１３）式の関係を満たすことが必要となる。
【００８７】
【数１３】

【００８８】
上記の（１３）式をｓについて解けば、線形光フィルタ１１Ａ’の振幅ｓは０．７４９７４と求まる。
【００８９】
図１１は、上記のようにして最適化が図られた線形光フィルタ１１Ａ’のフィルタ特性を示したものである。ただし、図１１の縦軸は線形光フィルタ１１Ａ’における減衰量を負の値で表している。また、このような線形光フィルタ１１Ａ’を通過して一次チルト成分の補償が行われたＷＤＭ信号光は、例えば図１２に示すようなスペクトルとなる。ここで、図１２のスペクトルについても、各光周波数ν_ｉに対応した光パワーＰ_ｉをそれぞれ読み取り、上記の場合と同様の手順に従って一次チルト成分を演算してみると、図１３に示すように傾きＤが略０となっており、第１補償部１１’によって一次チルト成分が効率的に補償されていることが分かる。また、図１０および図１３を比較すると明らかなように、信号光周波数帯域の両端における信号光レベルの偏差も格段に小さく制御できていることが分かる。
【００９０】
上記のようなアルゴリズムに従って第１補償部１１’のフィルタ特性の最適化が完了すると、次に、第２補償部１２’についてのフィルタ特性の最適化が行われる。
【００９１】
具体的には、第１補償部１１’を通過して一次チルト成分が補償されたＷＤＭ信号光が、第２補償部１２’を通過し（この時のフィルタ特性は光周波数に対して平坦な状態）、その一部が光分岐部１４でモニタ光として分岐されて光モニタ部１５に送られる。光モニタ部１５では、上記の図１２に示したような一次チルト成分補償後のＷＤＭ信号光のスペクトルが検出され、各々の光周波数ν_ｉに対応した光パワーＰ_ｉおよび光信号対雑音比ＯＳＮＲ_ｉが制御部１６に伝えられる。
【００９２】
制御部１６では、光モニタ部１５でのモニタ結果を基に、一次チルト成分補償後の信号光パワーの光周波数特性に関するフーリエ級数展開が行われ、フーリエフィルタ１２Ａ’を構成する複数の周期フィルタそれぞれの振幅および位相が最適化される。このフーリエ級数展開法に基づいた第２補償部１２’の最適化のアルゴリズムは、例えば特開２０００−１９９８８０号公報等に記載された設計手法を適用することが可能であり、ここではその概略について簡単に説明する。
【００９３】
フーリエ級数展開法を用いたフィルタ設計では、フーリエフィルタ１２Ａ’を構成する複数の周期フィルタのうちで最も長周期のフィルタのＦＳＲを１とすると、他の周期フィルタのＦＳＲは１／ｎ（ただし、ｎは整数とする）となるように設計される。このとき、最も長周期のフィルタのＦＳＲは、利得等化を行う信号光周波数帯域に余裕周波数帯域を足し合わせて設計される。従来の技術では、通常、余裕周波数帯域は信号光周波数帯域の約半分程度に設定される。しかし、本利得等化器１ｂにおいては、第１利得等化部１１’により信号光周波数帯域の両端部分における信号光レベルの偏差が補償されているため、余裕周波数帯域を非常に狭くすることができる。これにより、フーリエフィルタ１２Ａ’の段数の低減を図ることが可能になる。
【００９４】
具体的に、前述の図１２や図１３に示したような一次チルト成分補償後の信号光パワーの光周波数特性について考えると、信号光周波数帯域は３．９５ＴＨｚであり、この信号光周波数帯域に少しの余裕周波数帯域を設け、最も長周期のフィルタのＦＳＲを５．１ＴＨｚとすることが可能である。この５．１ＴＨｚを基本周期として、他の周期フィルタのＦＳＲを、例えば、２．５５ＴＨｚ、１．７ＴＨｚ、１．２７５ＴＨｚ、１．０２ＴＨｚ、０．８５ＴＨｚおよび０．７２９ＴＨｚに設定して合計７段の周期フィルタの組み合わせによりフーリエフィルタ１２Ａ’を構成する。７段の周期フィルタの各振幅は、信号光パワーの光周波数特性をフーリエ級数展開した結果に応じて、例えば図１４に示すようにそれぞれ個別に設定される。このようにしてＦＳＲおよび振幅の設定された７段の周期フィルタを組み合わせることによって、フーリエフィルタ１２Ａ’のフィルタ特性は、図１５に示すように７段のフィルタの和に相当する形状に最適化される。なお、ここでは７段構成のフーリエフィルタ１２Ａ’を例示したが、本発明におけるフーリエフィルタの段数はこれに限らず、要求される補償精度に応じて適宜に設定することが可能である。
【００９５】
上記のようにしてフィルタ特性が最適化された第２補償部１２’を通過してうねり成分が補償されたＷＤＭ信号光は、例えば図１６に示すようなスペクトルとなる。また、図１６に示すＷＤＭ信号光のスペクトルについて、各光周波数ν_ｉに対応した光パワーＰ_ｉをそれぞれ読み取ってプロットすると図１７のようになる。図１６および図１７から明らかなように、信号光パワーの光周波数間における偏差が非常に高い精度で補償されており、信号光周波数帯域の全体に亘って平坦な信号光パワーを持つＷＤＭ信号光が得られることが分かる。
【００９６】
以上のようにして起動時における第１補償部１１’および第２補償部１２’の各フィルタ特性の最適化が完了した後は、光モニタ部１５でのモニタ結果が制御部１６に随時フィードバックされ、光増幅器の動作状態の変化やＷＤＭ信号光の使用される光周波数の変化などに応じて、起動時と同様のアルゴリズムに従い第１補償部１１’および第２補償部１２’の各フィルタ特性が最適化される。
【００９７】
このように第２実施形態の利得等化器１ｂによれば、上述した第１実施形態の場合と同様の効果が得られるのに加えて、第１補償部１１’および第２補償部１２’の各フィルタ特性を可変にしてフィードバック制御するようにしたことで、光増幅器の利得波長特性の補償をより高い精度で安定して行うことが可能になる。
【００９８】
なお、上述した第１、２実施形態の利得等化器１ａ，１ｂでは、光増幅器の利得波長特性の一次チルト成分を補償する第１補償部１１，１１’を入力側に配置し、うねり成分を補償する第２補償部１２，１２’を出力側に配置してカスケード接続するようにしたが、第１、２補償部の配置を入れ替えて、第２補償部１２，１２’を入力側に配置し、第１補償部１１，１１’を出力側に配置してカスケード接続することも可能である。
【００９９】
次に、上述したような第１、２実施形態の利得等化器１ａ，１ｂを用いて構成した光増幅器について説明する。
【０１００】
図１８は、本発明の第３実施形態による光増幅器の構成を示すブロック図である。
【０１０１】
図１８において、本光増幅器２ａは、例えば、段の光増幅部２１，２２をカスケード接続して光出力の増大を図るようにした公知の構成の光増幅器について、光増幅部２１，２２の段間に上述した第１実施形態の利得等化器１ａを適用したものである。
【０１０２】
この光増幅器２ａは、具体的には、入力端子ＩＮと出力端子ＯＵＴの間に順にカスケード接続した、光増幅部２１、利得等化器（ＧＥＱ）１Ａ、可変光減衰器（ＶＯＡ）２３および光増幅部２２と、光増幅部２１への入力光パワーをモニタするための光分岐部２４Ａおよび受光器（ＰＤ）２５Ａと、光増幅部２１からの出力光パワーをモニタするための光分岐部２４Ｂおよび受光器２５Ｂと、光増幅部２２への入力光パワーをモニタするための光分岐部２４Ｃおよび受光器２５Ｃと、光増幅部２２からの出力光パワーをモニタするための光分岐部２４Ｄおよび受光器２５Ｄと、各光分岐部２４Ａ〜２４Ｄでのモニタ結果に基づいて光増幅部２１，２２および可変光減衰器２３の動作状態を制御する制御回路２６とを備える。
【０１０３】
制御回路２６は、光増幅器２ａへの入力光パワーが変動したときでも、各光増幅部２１，２２の利得波長特性が変化しないように、各々の光増幅部２１，２２の利得を一定に制御するＡＧＣを行うと共に、後段の光増幅部２２に対する入力パワー（および出力パワー）が一定となるように、可変光減衰器２３における減衰量を制御して入力ダイナミックレンジの確保を行う。なお、上記のような光増幅器の制御方式は、従来の光増幅器に適用されている制御方式と同様であるので、ここでの説明を省略する。
【０１０４】
上記のような構成の光増幅器２ａでは、前段および後段の光増幅部２１，２２の各利得波長特性が、段間に挿入された本発明による利得等化器１ａによって高い精度で補償されるようになる。これにより、本光増幅器２ａは、入力端子ＩＮに入力されるＷＤＭ信号光を所望のレベルまで増幅して、波長間における信号光レベルの揃ったＷＤＭ信号光を出力端子ＯＵＴから出力することができる。
【０１０５】
次に、本発明の第４実施形態による光増幅器について説明する。
【０１０６】
図１９は、第４実施形態の光増幅器の構成を示すブロック図である。
【０１０７】
図１９において、本光増幅器２ｂは、例えば、前述したような公知の２段増幅構成の光増幅器について、上述の第２実施形態で説明したフィードバック制御ありの利得等化器１ｂを設けたものである。
【０１０８】
この光増幅器２ｂは、具体的には、前段および後段の光増幅部２１，２２の段間に、前述の第２実施形態で説明した可変のフィルタ特性を持つ第１補償部１１’および第２補償部１２’をカスケード接続し、後段の光増幅部２２から出力される増幅光の一部を光分岐部２４Ｄ，１４で分岐して光モニタ部１５に送り、光モニタ部１５でのモニタ結果に基づいて、制御回路２６’により第１補償部１１’および第２補償部１２’の各フィルタ特性をフィードバック制御するようにしたものである。
【０１０９】
上記の光増幅器２ｂでは、フィードバック制御が行われる利得等化器１ｂを適用したことにより、各光増幅部２１，２２の利得波長特性を広い範囲に亘ってダイナミックに補償できるため、前述した第３実施形態の場合のように各光増幅部２１，２２に対するＡＧＣを行う必要はなくなり、各光増幅部２１，２２の出力を一定に制御するＡＰＣを行えばよい。このため、第３実施形態において各光増幅部２１，２２の前段に配置していた光分岐部２４Ａ，２４Ｃおよび受光器２５Ａ，２５Ｃは、本実施形態では省略することが可能となる。また、後段の光増幅部２２への入力パワーを一定にする必要もなくなるため、第３実施形態において各光増幅部２１，２２の間に配置していた可変光減衰器２３についても、本実施形態では省略することが可能となる。
【０１１０】
このように第４実施形態の光増幅器によれば、第１、２補償部１１’，１２’の各フィルタ特性をフィードバック制御する利得等化器１ｂを適用したことによって、より簡素な構成で安定した光増幅を行うことが可能になる。
【０１１１】
なお、上述した第３、４実施形態では、前段および後段の光増幅部２１，２２の間に第１、２補償部を配置した構成について説明したが、本発明の光増幅器の構成はこれに限られるものではない。例えば図２０に示す光増幅器２ｃのように、入力端子ＩＮと前段の光増幅部２１との間に第１、２補償部１１’，１２’を配置してもよい。また、例えば図２１に示す光増幅器２ｄのように、後段の光増幅部２２と光分岐部２４Ｄの間に第１、２補償部１１’，１２’を配置することも可能である。ただし、レベルダイヤやＯＳＮＲ劣化などの点を考慮すると、第１、２補償部１１’，１２’の挿入位置は、図１９に示した光増幅器２ｂのように、各光増幅部２１，２２の間とするのが望ましい。
【０１１２】
次に、上述したような利得等化器１ａ，１ｂまたは光増幅器２ａ〜２ｄを用いて構成したＷＤＭ光伝送システムについて説明する。
【０１１３】
図２２は、本発明の第５実施形態によるＷＤＭ光伝送システムの構成を示すブロック図である。
【０１１４】
図２２に示すＷＤＭ光伝送システム３ａは、例えば、ＷＤＭ信号光を送受信する光送信装置３１および光受信装置３２の間を光伝送路３３で接続すると共に、該光伝送路３３上に所要の間隔で光増幅器３４を配置してＷＤＭ信号光を中継伝送する公知のシステム構成について、各中継区間の光増幅器３４にそれぞれ対応させて上述の第１、２実施形態に示した利得等化器１ａまたは１ｂ（図中はＧＥＱ１として示す）を配置したものである。
【０１１５】
このようなＷＤＭ光伝送システム３ａでは、各中継区間の光増幅器３４の利得波長特性が各々に対応した利得等化器（ＧＥＱ）１によって高い精度で補償されるようになるため、光送信装置３１および光受信装置３２間におけるＷＤＭ信号光の中継伝送を確実に安定して行うことが可能になる。
【０１１６】
なお、上記第５実施形態のＷＤＭ光伝送システムでは、一般的な光増幅器３４に対応させて本発明の利得等化器を付設する構成としたが、公知のシステム構成における各中継区間の光増幅器３４として、例えば図２３に示すＷＤＭ光伝送システム３ｂのように上述の第３、４実施形態に示した光増幅器２ａ〜２ｄ（図中は光増幅器２として示す）のいずれかを適用するようにしてもよい。
【０１１７】
また、各々の中継区間ごとに利得等化器を配置して利得等化を行う一例を示したが、２つ以上の中継区間についての利得等化を１つの利得等化器によりまとめて行うことも可能である。この場合、例えば図２４に示すＷＤＭ光伝送システム３ｃのように、第１補償部１１’と第２補償部１２’を異なる中継区間に配置し、すなわち、第１補償部１１’および第２補償部１２’の間に光増幅器３４および光伝送路３３が存在するような配置として、異なる中継区間の第１、２補償部１１’，１２’の各フィルタ特性を制御部１６により連動してフィードバック制御することも可能である。
【０１１８】
以上、本明細書で開示した主な発明について以下にまとめる。
【０１１９】
（付記１） 光の波長特性を補償する利得等化器であって、
補償対象波長領域について単調増加または単調減少する損失波長特性を有し、前記補償対象波長領域内の波長特性の直線的な利得成分を補償する第１補償部と、
該第１補償部にカスケード接続する複数の周期フィルタの組み合わせにより、前記第１補償部で未補償となる前記補償対象波長領域内の波長特性を補償する第２補償部と、
を備えて構成されることを特徴とする利得等化器。
【０１２０】
（付記２） 付記１に記載の利得等化器であって、
前記第１補償部および前記第２補償部が、可変の損失波長特性をそれぞれ有すると共に、
前記第１補償部および前記第２補償部を通過した光をモニタする光モニタ部と、
該光モニタ部のモニタ結果に基づいて前記補償対象波長領域内の波長特性を判断し、該判断結果に応じて前記第１補償部および前記第２補償部の各損失波長特性をフィードバック制御する制御部と、
を備えて構成されることを特徴とする利得等化器。
【０１２１】
（付記３） 付記１に記載の利得等化器であって、
前記第１補償部は、前記補償対象波長領域よりも広い自由スペクトル領域を持ち、周期的な損失波長特性が前記補償対象波長領域において単調増加または単調減少する長周期フィルタを備えることを特徴とする利得等化器。
【０１２２】
（付記４） 付記１に記載の利得等化器であって、
前記第２補償部は、自由スペクトル領域が互いに異なる複数の周期フィルタを組み合わせて構成した光フィルタを備えることを特徴とする利得等化器。
【０１２３】
（付記５） 付記２に記載の利得等化器であって、
前記制御部は、前記光モニタ部のモニタ結果について最小二乗線形フィッティング法を適用して前記補償対象波長領域内の波長特性の一次傾斜値を計算し、前記第１補償部の損失波長特性が前記一次傾斜値に対応した割合で単調増加または単調減少するように、前記第１補償部に対するフィードバック制御を行うことを特徴とする利得等化器。
【０１２４】
（付記６） 付記２に記載の利得等化器であって、
前記制御部は、前記光モニタ部のモニタ結果についてフーリエ級数展開法を適用し、前記第２補償部の各周期フィルタの振幅および位相がフーリエ級数展開の演算結果に対応するように、前記第２補償部に対するフィードバック制御を行うことを特徴とする利得等化器。
【０１２５】
（付記７） 付記２に記載の利得等化器であって、
前記光モニタ部は、波長多重信号光の各波長に対応した信号光パワーをモニタし、
前記制御部は、前記光モニタ部でモニタされる各波長の信号光パワーを用いた演算処理により前記補償対象波長領域内の波長特性を判断することを特徴とする利得等化器。
【０１２６】
（付記８） 付記２に記載の利得等化器であって、
前記光モニタ部は、波長多重信号光の各波長に対応した光信号対雑音比をモニタし、
前記制御部は、前記光モニタ部でモニタされる各波長の光信号対雑音比を用いた演算処理により前記補償対象波長領域内の波長特性を判断することを特徴とする利得等化器。
【０１２７】
（付記９） 波長多重信号光を一括して増幅する光増幅部と、
補償対象波長領域について単調増加または単調減少する損失波長特性を有し、前記光増幅部の利得波長特性の直線的な利得成分を補償する第１補償部と、
該第１補償部にカスケード接続する複数の周期フィルタの組み合わせにより、前記第１補償部で未補償となる前記光増幅部の利得波長特性を補償する第２補償部と、
を備えて構成されることを特徴とする光増幅器。
【０１２８】
（付記１０） 付記９に記載の光増幅器であって、
前記第１補償部および前記第２補償部が、可変の損失波長特性をそれぞれ有すると共に、
前記第１補償部および前記第２補償部を通過した波長多重信号光をモニタする光モニタ部と、
該光モニタ部のモニタ結果に基づいて前記光増幅部の利得波長特性を判断し、該判断結果に応じて前記第１補償部および前記第２補償部の各損失波長特性をフィードバック制御する制御部と、
を備えて構成されることを特徴とする光増幅器。
【０１２９】
（付記１１） 付記９に記載の光増幅器であって、
前記第１補償部および前記第２補償部は、カスケード接続された複数の光増幅部の各利得波長特性を補償することを特徴とする光増幅器。
【０１３０】
（付記１２） 付記１１に記載の光増幅器であって、
前記第１補償部および前記第２補償部は、前記カスケード接続された複数の光増幅部の間に配置されることを特徴とする光増幅器。
【０１３１】
（付記１３） 光伝送路上に配置された光増幅器により波長多重信号光を一括して増幅すると共に、前記光増幅器の利得波長特性を利得等化器により補償して中継伝送する光伝送システムであって、
前記利得等化器は、補償対象波長領域について単調増加または単調減少する損失波長特性を有し、前記光増幅器の利得波長特性の直線的な利得成分を補償する第１補償部と、該第１補償部にカスケード接続する複数の周期フィルタの組み合わせにより、前記第１補償部で未補償となる前記光増幅器の利得波長特性を補償する第２補償部と、を備えることを特徴とする光伝送システム。
【０１３２】
（付記１４） 付記１３に記載の光伝送システムであって、
前記利得等化器は、前記第１補償部および前記第２補償部が、可変の損失波長特性をそれぞれ有すると共に、
前記第１補償部および前記第２補償部を通過した波長多重信号光をモニタする光モニタ部と、
該光モニタ部のモニタ結果に基づいて前記光増幅器の利得波長特性を判断し、該判断結果に応じて前記第１補償部および前記第２補償部の各損失波長特性をフィードバック制御する制御部と、
を備えて構成されたことを特徴とする光伝送システム。
【０１３３】
（付記１５） 付記１４に記載の光伝送システムであって、
前記光伝送路上に複数の光増幅器が配置され、波長多重信号光が複数の中継区間を伝送されるとき、
前記利得等化器は、各中継区間にそれぞれ配置されることを特徴とする光伝送システム。
【０１３４】
（付記１６） 付記１４に記載の光伝送システムであって、
前記光伝送路上に複数の光増幅器が配置され、波長多重信号光が複数の中継区間を伝送されるとき、
前記利得等化器は、前記第１補償部および前記第２補償部が異なる中継区間に配置されることを特徴とする光伝送システム。
【０１３５】
（付記１７） 光の波長特性を補償する利得等化方法であって、
補償対象波長領域について単調増加または単調減少する損失波長特性を有する第１補償部により、前記補償対象波長領域内の波長特性の直線的な利得成分を補償し、
前記第１補償部にカスケード接続する複数の周期フィルタの組み合わせからなる第２補償部により、前記第１補償部で未補償となる前記補償対象波長領域内の波長特性を補償することを特徴とする利得等化方法。
【０１３６】
（付記１８） 付記１７に記載の利得等化方法であって、
前記第１補償部および前記第２補償部を通過した光をモニタし、
該モニタ結果に基づいて前記補償対象波長領域内の波長特性を判断し、
該判断結果に応じて前記第１補償部および前記第２補償部の各損失波長特性をフィードバック制御することを特徴とする利得等化方法。
【０１３７】
【発明の効果】
以上説明したように、本発明の利得等化技術によれば、利得等化を行う光について、補償対象波長領域内の波長特性の直線的な利得成分を第１補償部で補償し、その第１補償部で補償されなかったうねり成分を第２補償部で補償するようにしたことで、利得等化を高い精度で効率的に行うことが可能になる。
【図面の簡単な説明】
【図１】本発明の第１実施形態による利得等化器の構成を示すブロック図である。
【図２】同上第１実施形態に用いられる線形光フィルタの具体的な構成の一例を示す図である。
【図３】同上第１実施形態において利得等化の対象となる光増幅器の利得波長特性の一例を示す図である。
【図４】同上第１実施形態において第１補償部で一次チルト成分が補償された後の利得波長特性を示す図である。
【図５】図４の利得波長特性に応じて設計したフーリエフィルタの損失波長特性を示す図である。
【図６】図３の利得波長特性に応じて設計したフーリエフィルタ（従来技術）の損失波長特性を示す図である。
【図７】本発明の第２実施形態による利得等化器の構成を示すブロック図である。
【図８】同上第２実施形態に用いられる線形光フィルタの具体的な構成の一例を示す図である。
【図９】同上第２実施形態において起動時にモニタされるＷＤＭ信号光のスペクトル例を示す図である。
【図１０】図９について光周波数に対応した光パワーを読み取り一次チルト成分を計算した結果を示す図である。
【図１１】図１０の結果に対応して最適化された線形光フィルタのフィルタ特性を示す図である。
【図１２】図１１の線形光フィルタを通過して一次チルト成分が補償されたＷＤＭ信号光のスペクトルを示す図である。
【図１３】図１２について光周波数に対応した光パワーを読み取り一次チルト成分を計算した結果を示す図である。
【図１４】上記第２実施形態においてフーリエフィルタを構成する複数の周期フィルタの特性例を示す図である。
【図１５】図１４の各周期フィルタの組み合わせによって得られるフーリエフィルタの特性を示す図である。
【図１６】図１５のフーリエフィルタを通過してうねり成分が補償されたＷＤＭ信号光のスペクトルを示す図である。
【図１７】図１６について光周波数に対応した光パワーを読み取った結果を示す図である。
【図１８】本発明の第３実施形態による光増幅器の構成を示すブロック図である。
【図１９】本発明の第４実施形態による光増幅器の構成を示すブロック図である。
【図２０】同上第４実施形態に関連する他の構成例を示すブロック図である。
【図２１】同上第４実施形態に関連する別の構成例を示すブロック図である。
【図２２】本発明の第５実施形態によるＷＤＭ光伝送システムの構成を示すブロック図である。
【図２３】同上第５実施形態に関連する他のシステム構成例を示すブロック図である。
【図２４】同上第５実施形態に関連する別のシステム構成例を示すブロック図である。
【図２５】従来の利得等化器に関する問題点を説明するための図である。
【図２６】従来の利得等化器に関する他の問題点を説明するための図である。
【符号の説明】
１Ａ，１Ｂ 利得等化器（ＧＥＱ）
１１，１１’ 第１補償部
１１Ａ 線形光フィルタ
１２，１２’ 第２補償部
１２Ａ フーリエフィルタ
１３ 光路
１４ 光分岐部
１５ 光モニタ部
１６ 制御部
２Ａ〜２Ｄ 光増幅器
２１，２２ 光増幅部
２３ 可変光減衰器（ＶＯＡ）
２４Ａ〜２４Ｄ 光分岐部
２５Ａ〜２５Ｄ 受光器（ＰＤ）
２６ 制御回路
ＩＮ 入力端子
ＯＵＴ 出力端子[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a gain equalization technique for compensating for wavelength characteristics of light, and more particularly, to a gain equalization technique for compensating for wavelength characteristics in a compensation target wavelength region by a combination of a plurality of periodic filters, and application thereof. The present invention relates to an optical amplifier and an optical transmission system.
[0002]
[Prior art]
In recent years, optical fiber communication systems have been put into practical use with the emergence of low-loss optical fibers due to improvements in optical fiber manufacturing technology. In addition, in order to realize long-distance optical transmission, an optical amplifier for amplifying signal light has been put to practical use as a technology for compensating for loss in an optical fiber transmission line, and a wavelength for increasing the transmission line capacity has been used as a technology for increasing the transmission line capacity. Multiplexed (WDM) optical transmission systems are also known. This WDM optical transmission method is a method of multiplexing and transmitting a plurality of signal lights having different wavelengths using one fiber.
[0003]
By the way, when using an optical amplifier that collectively amplifies multiplexed signal light, the signal light level after passing through the optical amplifier or the like depends on the wavelength due to the wavelength dependence (gain wavelength characteristic) of the gain of the optical amplifier or the like. Is known to be different. The deviation between the wavelengths of the signal light levels generated by the optical amplifiers and the like is accumulated in a system in which a plurality of optical amplifiers are cascaded. As described above, the wavelength dependence of the gain of an optical amplifier or the like causes a deviation of the signal light level and a deviation of the optical signal-to-noise ratio (OSNR) for each wavelength. This poses a problem in that it restricts and deteriorates the receiving sensitivity at the receiving end.
[0004]
Conventionally, various techniques for compensating for gain wavelength characteristics of an optical amplifier (gain equalization) have been proposed. For example, in the following Patent Document 1 and the like, a change period on the wavelength axis of the transmittance, that is, a gain equalizer configured by cascading a plurality of periodic filters having different free spectral regions (FSR) is used. Techniques for compensating for gain wavelength characteristics of optical amplifiers have been proposed. As is apparent from the Fourier series expansion method, this conventional technique is based on the principle that a filter characteristic of an arbitrary shape can be realized by superimposing a plurality of sine waves having different periods from each other on the wavelength axis (or A gain equalizer having a desired filter characteristic by cascading a plurality of periodic filters having different FSRs using a Mach-Zehnder type optical filter having a sinusoidal variable transmission characteristic on the optical frequency axis) as a periodic filter. Has been realized.
[0005]
In the following Patent Document 2 and the like, instead of configuring a gain equalizer by combining a plurality of periodic filters as described above, a maximum value of a loss is provided in a predetermined wavelength band, and the maximum value is determined. There has been proposed a technique for compensating for the gain wavelength characteristic of an optical amplifier using an optical filter whose loss wavelength characteristic is asymmetric about a given wavelength. In this prior art, the gain wavelength characteristic of the optical amplifier can be equalized over a wide wavelength band by one optical filter.
[0006]
[Patent Document 1]
JP-A-6-276154
[Patent Document 2]
JP 2000-354005 A
[0007]
[Problems to be solved by the invention]
However, the above-described related art has the following problems. That is, when equalizing the gain wavelength characteristic of an optical amplifier using a periodic filter, first, gain equalization is actually performed on a compensation target wavelength band (for example, a signal light wavelength band) to be subjected to gain equalization. A wavelength band to be performed is determined, and a periodic filter is designed in accordance with the wavelength band. Specifically, for example, as shown in FIG. _S Wavelength region Δλ required for both ends of _M Is set, and the signal light wavelength band Δλ is set according to the gain wavelength characteristic of the optical amplifier to be subjected to gain equalization. _S The loss (or transmission) wavelength characteristic for is designed. And the signal light wavelength band Δλ _S For other bands, the signal light wavelength band Δλ _S Are repeated on the wavelength axis, and the signal light wavelength band Δλ _S The characteristic at the boundary with the margin wavelength region Δλ _M The filter characteristics are designed so as to provide a smooth connection. More specifically, the signal light wavelength band Δλ _S And the marginal wavelength region Δλ _M By performing Fourier series expansion of the filter characteristics designed for the wavelength band obtained by adding the above, the characteristics of each periodic filter are designed. According to such a filter design, an optical filter having a periodic loss wavelength characteristic is realized, and the gain wavelength characteristic of the optical amplifier is compensated.
[0008]
In the above filter design, the marginal wavelength band Δλ _M Is too wide, the number of required periodic filters increases. Specifically, for example, the marginal wavelength band Δλ _M Is the signal light wavelength band Δλ _S When set to 0.5 times the number of required periodic filters, the margin wavelength band Δλ _M Is 1.5 times larger than the case where is not set. Therefore, the surplus wavelength band Δλ _M It is more efficient to set as small as possible. However, as illustrated in FIG. 25B, the marginal wavelength band Δλ _M If the wavelength band for actually performing gain equalization is set to be substantially the same as the signal light wavelength band, the signal light wavelength band Δλ _S (Specifically, the marginal wavelength region Δλ _M In (2), the filter characteristic in which the loss changes abruptly is designed as a target value, and there is a problem that it is difficult to accurately realize an optical filter having such a filter characteristic. Such a problem is caused by the signal light wavelength band Δλ as illustrated in FIG. _S Will be noticeable when designing an optical filter corresponding to the case where the gain deviation of the optical amplifier at both end portions is large.
[0009]
As described above, the signal light wavelength band Δλ _S The shape of the filter characteristic in is realized with the required accuracy by synthesizing a plurality of periodic filters having different FSRs based on the Fourier series expansion method. Also, the signal light wavelength band Δλ _S Has a signal light wavelength band Δλ _S It will be repeated in other areas. At this time, the signal light wavelength band Δλ _S If the deviation of the signal light power at both ends is large, the target filter shape becomes the signal light wavelength band Δλ. _S Is almost discontinuous at both ends, the error between the filter shape actually obtained by combining a plurality of periodic filters and the target filter shape increases. Specifically, for example, when the target filter shape is a saw-tooth waveform, the actually obtained filter shape becomes a smooth curve. When the gain wavelength characteristic of the optical amplifier is compensated using such an optical filter having a large error, the signal light wavelength band Δλ _S The compensation accuracy in the vicinity of both ends will be degraded.
[0010]
Also, for example, as shown in FIG. 26, when the amplitude of the loss wavelength characteristic 100 of each periodic filter synthesized based on the Fourier series expansion method becomes large, the combined filter characteristic 101 is sharp. The shape becomes steep. For this reason, there is also a problem that fine waviness (ripple) is generated as residue in the signal light power corresponding to each wavelength obtained after gain equalization.
[0011]
The present invention has been made by paying attention to the above problem, and it is possible to efficiently compensate the wavelength characteristics of the light to be subjected to gain equalization in the compensation target wavelength region with high accuracy by combining a plurality of periodic filters. It is an object of the present invention to realize a gain equalization technique and to provide an optical amplifier and an optical transmission system to which the technique is applied.
[0012]
[Means for Solving the Problems]
In order to achieve the above object, the gain equalizer of the present invention is a gain equalizer that compensates for optical wavelength characteristics, and has a loss wavelength characteristic that monotonically increases or monotonically decreases with respect to a wavelength region to be compensated. Due to the combination of the first compensator for compensating for the linear gain component of the wavelength characteristic in the compensation target wavelength region and a plurality of periodic filters cascaded to the first compensator, the first compensator is not compensated. And a second compensator for compensating for wavelength characteristics in the compensation target wavelength region. Here, “equalization” means that compensation is performed so that undesired wavelength characteristics of light become the same or approximate to the preferable wavelength characteristics.
[0013]
In the gain equalizer having such a configuration, since the light for gain equalization passes through the first compensator, the linear gain component of the wavelength characteristic in the wavelength region to be compensated is compensated. The wavelength characteristic in the wavelength region to be compensated which is not compensated by the above is compensated by the second compensator, and the wavelength characteristic of the light is equalized. At this time, since the linear gain component is compensated by the first compensating unit, the deviation of the signal light power at both ends of the compensation target wavelength region is reduced. Accordingly, the filter design of the second compensator configured by combining a plurality of periodic filters can be efficiently performed with high accuracy.
[0014]
Further, in the gain equalizer, the first compensator and the second compensator have variable loss wavelength characteristics, respectively, and the light that has passed through the first compensator and the second compensator is An optical monitoring unit for monitoring, and a wavelength characteristic in the wavelength region to be compensated is determined based on a monitoring result of the optical monitoring unit. Each loss of the first compensation unit and the second compensation unit is determined according to the determination result. It is preferable that the control unit includes a control unit that performs feedback control of the wavelength characteristic.
[0015]
According to this configuration, the variable loss wavelength characteristics of the first and second compensators are feedback-controlled according to the state of the light that has passed through the first and second compensators and gain-equalized. It is possible to stably equalize the wavelength characteristics in the wavelength region with higher accuracy.
[0016]
Further, as a specific mode of the gain equalizer, the control unit applies a least-squares linear fitting method to a monitoring result of the optical monitoring unit to apply a primary slope value of a wavelength characteristic in the compensation target wavelength region. And the feedback control to the first compensator may be performed such that the loss wavelength characteristic of the first compensator monotonically increases or decreases at a rate corresponding to the primary slope value. Accordingly, the linear gain component of the wavelength characteristic in the compensation target wavelength region is more reliably compensated for in the first compensation unit.
[0017]
The optical amplifier and the optical transmission system according to the present invention are configured to compensate for the gain wavelength characteristic by applying the configuration of the gain equalizer as described above. Will be described in detail.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0019]
FIG. 1 is a block diagram showing a configuration of the gain equalizer according to the first embodiment of the present invention.
[0020]
In FIG. 1, the gain equalizer 1a according to the first embodiment includes, for example, a first compensator 11, a second compensator 12, and an optical path 13, and the first compensator 11 and the second compensator 12 receive input signals. A cascade connection is made between the terminal IN and the output terminal OUT via the optical path 13.
[0021]
The first compensator 11 compensates, for example, a primary tilt component corresponding to a linear gain component of the gain wavelength characteristic of an optical amplifier (not shown) connected to the input terminal IN or the output terminal OUT. Here, the primary tilt means a slope of the gain when the gain characteristic with respect to the wavelength of the optical amplifier is regarded as a straight line. The compensation of the primary tilt component of the gain wavelength characteristic means that a gain deviation at both ends of a compensation target wavelength band (for example, a signal light wavelength band) is compensated. Therefore, in other words, the first compensator 11 has a wavelength characteristic that changes linearly in the band for signal transmission corresponding to the gain wavelength characteristic of the optical amplifier, and the slope of the wavelength characteristic is appropriately designed. It was done.
[0022]
Specifically, the first compensator 11 includes an optical filter (hereinafter, referred to as a linear optical filter) 11A whose change in loss (or transmittance) with respect to wavelength monotonically increases or decreases with respect to the signal light wavelength band. . As the linear optical filter 11A, for example, a long-period filter such as a Fabry-Perot etalon interferometer type filter or a fiber Bragg grating filter, or a dielectric multilayer filter or an acousto-optic filter can be used. When a long-period filter is used as the linear optical filter 11A, the long-period filter has a free spectrum region (FSR) wider than the signal light wavelength band, and a region where the periodic loss wavelength characteristic increases or decreases monotonically. Are used corresponding to the signal light wavelength band.
[0023]
As an example of a specific configuration of the linear optical filter 11A, FIG. 2 shows a configuration of a Fabry-Perot etalon interferometer type filter. In the configuration example of FIG. 2, the signal light emitted from one end of the optical path on the input side is sent to the Fabry-Perot etalon interferometer type filter 11C via the lens 11B and transmitted through the Fabry-Perot etalon interferometer type filter 11C. Is focused on one end of the optical path on the output side via the lens 11D.
[0024]
The second compensator 12 receives the signal light whose primary tilt component has been compensated in the first compensator 11 via the optical path 13 and gain wavelength characteristics of the optical amplifier that has not been compensated in the first compensator 11. In other words, it compensates for the primary tilt component and the secondary or higher-order tilt component (hereinafter referred to as a swell component) that could not be completely compensated by the first compensating unit 11. The second compensator 12 includes an optical filter (hereinafter, referred to as a Fourier filter) 12A configured by combining a plurality of periodic filters based on a Fourier series expansion method in the same manner as in the related art. In the Fourier filter 12A, the free spectral range (FSR) of a plurality of periodic filters combined is 1 / n times the fundamental period determined according to the gain wavelength characteristic in the signal light wavelength band of the optical amplifier. Here, n is an integer). As each of the periodic filters constituting the Fourier filter 12A, for example, a Fabry-Perot etalon interferometer type filter or a fiber Bragg grating filter as shown in FIG. 2 can be used.
[0025]
The optical path 13 is provided between the input terminal IN of the gain equalizer 1a and the input terminal of the first compensator 11, between the output terminal of the first compensator 11 and the input terminal of the second compensator 12, The input terminal of the compensator 12 and the output terminal OUT of the gain equalizer 1a are connected to each other using, for example, an optical fiber or an optical waveguide.
[0026]
Here, the configuration example in which the first compensating unit 11 and the second compensating unit are individually formed and connected to each other by the optical path 13 has been described. For example, the first compensating unit 11 and the second compensating unit are It is also possible to integrate them into a single device.
[0027]
Here, each filter characteristic of the first compensator 11 and the second compensator 12 will be described in detail.
[0028]
FIG. 3 is a diagram illustrating an example of a gain wavelength characteristic (output ASE spectrum) of an optical amplifier to be subjected to gain equalization. With respect to the gain wavelength characteristic shown in FIG. 3, the gain equalizer 1a compensates for the loss of the linear filter 11A of the first compensator 11 so that the first-order tilt component indicated by the dashed line in the figure is compensated. Characteristics are designed. Specifically, for the gain wavelength characteristic of FIG. 3, a first-order slope value is obtained by applying a known operation process such as a least-squares linear fitting method, and the first-order tilt component of the gain wavelength characteristic is canceled according to the result. The slope of the loss wavelength characteristic of the linear filter 11A (the rate at which the loss monotonically increases or monotonously decreases with respect to the change in wavelength) is determined. The details of how to determine the primary slope value by the least squares linear fitting method will be described later.
[0029]
When the light having the spectrum as shown in FIG. 3 is input to the first compensating unit 11 which determines the slope of the loss wavelength characteristic of the linear filter 11A as described above, the first compensating unit 11 The light output from has a spectrum as shown in FIG. 4, for example. The spectrum of the output light from the first compensator 11 corresponds to the undulation component of the gain wavelength characteristic of the optical amplifier, and the Fourier transform of the second compensator 12 is performed so that the undulation component of the gain wavelength characteristic is compensated. The loss wavelength characteristic of the filter 12A is designed. Specifically, the fundamental period is designed by adding a desired margin wavelength region to the signal light wavelength band of FIG. A required number of periodic filters having an FSR of 1 / n times (n = 1, 2, 3,...) Are designed by applying the Fourier series expansion method to the loss wavelength characteristic in the fundamental period. The details of how to obtain the filter characteristics by the Fourier series expansion method will be described later.
[0030]
When the loss wavelength characteristic of the Fourier filter 12A with respect to the signal light wavelength band is designed as described above, the Fourier filter 12A is realized using the characteristic as a target value. At this time, the Fourier filter 12A actually configured is a composite of the respective characteristics of the plurality of periodic filters, so that the target loss wavelength characteristic is repeated not only within the signal light wavelength band but also outside the signal light wavelength band. Will be. For this reason, as shown in FIG. 25 described above, if the loss wavelength characteristics are not smoothly connected at both end portions (margin wavelength region) of the signal light wavelength band, an error with respect to the target value increases. Since the primary tilt component is compensated by the unit 11, even if the margin wavelength region is set narrow, the loss wavelength characteristics can be smoothly connected.
[0031]
FIG. 5 shows the loss wavelength characteristics of the Fourier filter 12A designed according to the gain wavelength characteristics shown in FIG. However, in FIG. 5, about 1.15 nm is set as the minimum value of the surplus wavelength region in which the characteristics after compensating the primary tilt component can be smoothly connected, and the characteristics in the surplus wavelength region are linear for simplicity. Is approximated. Although the loss wavelength characteristics of three periods within the signal light wavelength band and outside the signal light wavelength band on the shorter wavelength side are shown here, the filter of the signal light wavelength band designed as the target value is shown. The characteristics are repeated for other wavelength regions. On the other hand, FIG. 6 shows a Fourier filter realized according to the target value designed according to the gain wavelength characteristic shown in FIG. 3, that is, the design is performed without compensating the primary tilt component as in the conventional case. 4 shows the loss wavelength characteristics of the Fourier filter when the wavelength is changed. However, the setting of the margin wavelength region and the like is set to the same condition as the case of FIG.
[0032]
As is clear from the comparison between FIG. 5 and FIG. 6, the loss wavelength characteristic of the Fourier filter 12A used in the second compensator 12 is marginal because the first compensator 11 compensates for the primary tilt component. Since the connection is very smooth in the wavelength region, it is possible to accurately match the actually obtained filter characteristic with the target value for which the filter is designed.
[0033]
In the gain equalizer 1a in which the first compensator 11 and the second compensator 12 having the above filter characteristics are cascaded via the optical path 13, for example, the WDM signal light amplified by the optical amplifier is input. By being input to the terminal IN and being sent to the first compensating unit 11 and the second compensating unit 12 in order, the deviation occurring between the signal light levels of the respective wavelengths is compensated, and each signal light level is flattened. The WDM signal light is output from the output terminal OUT. Further, for example, the WDM signal light before being amplified by the optical amplifier is input to the input terminal IN and sequentially sent to the first compensating unit 11 and the second compensating unit 12, so that the signal light level that can be generated in the optical amplifier is reduced. The WDM signal light whose wavelength deviation has been compensated in advance is output from the output terminal OUT.
[0034]
As described above, according to the gain equalizer of the first embodiment, the first compensator 11 that compensates the primary tilt component of the gain wavelength characteristic of the optical amplifier using the linear optical filter 11A, and the swell component is the Fourier filter 12A By cascade-connecting the second compensating unit for compensating by using, the gain wavelength characteristic of the optical amplifier can be reliably compensated with high accuracy. Further, in the present gain equalizer, the margin wavelength region can be set very narrow as compared with the conventional case, so that the number of periodic filters (the number of filter stages) combined to configure the Fourier filter 12A can be reduced. . As a result, efficient filter design can be performed.
[0035]
Next, a gain equalizer according to a second embodiment of the present invention will be described.
[0036]
FIG. 7 is a block diagram illustrating the configuration of the gain equalizer according to the second embodiment. The same parts as those in the configuration of the first embodiment shown in FIG. 1 are denoted by the same reference numerals, and the same applies to other drawings.
[0037]
7, a gain equalizer 1b according to the second embodiment uses, for example, a linear optical filter 11A ′ that can change the loss wavelength characteristic in the configuration of the first embodiment illustrated in FIG. A first compensator 11 'and a second compensator 12' using a Fourier filter 12A 'capable of changing a loss wavelength characteristic are applied, and the first compensator 11' and the second compensator 12 'are used. An optical branching unit 14, an optical monitoring unit 15, and a control unit 16 for performing feedback control of each filter characteristic are provided.
[0038]
The linear optical filter 11A ′ is an optical filter whose loss (or transmission) wavelength characteristic increases or decreases monotonically in the signal light wavelength band, as in the first embodiment, and the ratio of the monotonically increasing or monotonically decreasing loss. Is variably controlled. Examples of the linear optical filter 11A ′ having such variable characteristics include, for example, a Mach-Zehnder interferometer-type filter known in Japanese Patent Application Laid-Open No. 6-276154 and an etalon-type filter having a reflection film formed on the input / output surface of a magneto-optical crystal. A long-period filter such as a filter (for example, Japanese Patent Application No. 2002-146643) can be used.
[0039]
As shown in FIG. 8, for example, the Mach-Zehnder interferometer type filter has a configuration in which the input and output sides of two optical waveguides 11S and 11T having different lengths are connected by directional couplers 11U and 11V, respectively. By adjusting the bias current to the electrodes 11X and 11Y provided in the directional couplers 11U and 11V, it is possible to change the coupling efficiency and thereby change the amplitude of the filter. , 11V, the phase of the filter can be changed by adjusting the bias current to the electrode 11Z provided on the optical waveguide 11S.
[0040]
The above etalon-type filter forms a etalon-type optical filter by forming a reflection film on an input / output surface of a polarization variable controller using a magneto-optical effect such as a Faraday rotator, and controls the Faraday rotation angle. With this, the transmittance (or loss) can be changed. In addition, by forming the Faraday rotator with a pair of wedge-shaped movable magneto-optical crystals and controlling the resonator length of the etalon, it is possible to control the transmission characteristics on the wavelength axis.
[0041]
Further, as another specific example of the linear optical filter 11A 'having variable characteristics, an optical filter described in JP-A-11-271700 or the like can be used. This optical filter is specifically configured to include a polarization separation unit, a polarization variable unit, and a polarization wavelength characteristic changing element. First, input signal lights are orthogonalized by a polarization separation unit using a polarization beam splitter or the like. The light is separated into two lights, and the polarization state of one light is rotated by 90 degrees, so that the two lights have the same polarization state. Then, the polarization state of each light is changed by a polarization variable part utilizing a magneto-optical effect such as a Faraday rotator, and the polarization wavelength characteristic changing element such as a dielectric multilayer filter in which the transmission wavelength characteristic changes according to the polarization state. Each light is made incident, and each light that has passed through the polarization wavelength characteristic changing element is output along a path reverse to that before the light passed. In such an optical filter, the transmission wavelength characteristic can be variably controlled by controlling the amount of change in the polarization state in the polarization variable section. However, in the optical filter having the above-described configuration, only the rate of change in either the monotone increase or the monotone decrease can be controlled. Must be connected in cascade in two stages.
[0042]
The Fourier filter 12A ′ is an optical filter configured by combining a plurality of periodic filters based on the Fourier series expansion method, as in the first embodiment, and includes at least one or more periodic filters (preferably all The wavelength characteristic of the periodic filter can be variably controlled. As the periodic filter having such variable characteristics, for example, a Mach-Zehnder interferometer type filter or an etalon type filter having a reflection film formed on the input / output surface of the magneto-optical crystal as described above can be used. is there.
[0043]
The optical branching unit 14 is provided, for example, on the optical path 13 between the second compensating unit 12 and the output terminal OUT, and branches a part of the WDM signal light passing through the second compensating unit 12 as monitor light to perform optical monitoring. It is sent to the unit 15.
[0044]
The optical monitoring unit 15 detects signal light power or optical signal-to-noise ratio (OSNR) corresponding to each wavelength of the monitor light from the optical branching unit 14 and outputs a signal indicating the detection result to the control unit 16. Is what you do.
[0045]
The control unit 16 controls the variable filters of the first compensation unit 11 ′ and the second compensation unit 12 ′ based on the output signal from the optical monitoring unit 15 so that the gain wavelength characteristic of the optical amplifier can be compensated in an optimal state. The characteristics are respectively feedback controlled.
[0046]
Next, a specific algorithm of the feedback control by the control unit 16 will be described. However, it is assumed here that the WDM signal light amplified by the optical amplifier and having a deviation in the signal light level of each wavelength is input to the gain equalizer 1b. Specifically, as the input WDM signal light, a WDM signal light obtained by multiplexing 80 signal lights in a so-called C band region is considered, and the optical frequency (THz) and the optical power (THz) of the 80 wave signal lights are considered. dBm) and the optical signal-to-noise ratio (dB) _i , P _i And OSNR _i (Where i = 1 to 80). In the following description, a specific algorithm of feedback control is described using the optical frequency as one parameter, but it is of course possible to replace the optical frequency with a wavelength.
[0047]
First, at the time of starting the gain equalizer 1b, the first compensator 11 'and the second compensator 12' are set in a non-control state, so that each filter characteristic becomes flat with respect to the optical frequency. Initialized. In such a state, the WDM signal light, which has been amplified by the optical amplifier and has a deviation in the signal light level of each wavelength, has the first compensating unit 11 'and the second compensating unit 12 with almost no change in the optical frequency characteristics. , A part of the WDM signal light output from the second compensator 12 ′ is split as monitor light by the optical splitter 14, and the optical frequency ν of the monitor light _i Optical power P corresponding to _i And optical signal to noise ratio OSNR _i Is detected by the optical monitor unit 15 and transmitted to the control unit 16.
[0048]
FIG. 9 shows an example of the spectrum of the 80-wave WDM signal light to be monitored at the time of starting as described above. In the control unit 16 having received the monitoring result of the optical monitoring unit 15, first, the filter characteristics of the first compensating unit 11 'are optimized according to the following algorithm.
[0049]
In order to optimize the filter characteristics of the first compensator 11 ', first, the primary tilt component of the optical amplifier is calculated based on the monitoring result of the optical monitor 15. The primary tilt component of the optical amplifier is the optical frequency ν _i And optical power P _i Or the optical frequency ν _i And optical signal to noise ratio OSNR _i The constants C and D can be determined and determined by the least-squares linear fitting method, assuming that the relationship with follows a linear equation: y = C + D × x. This calculation can be handled very simply, for example, by using the following determinant:
[0050]
Optical power P for the above linear equation: y = C + D × x _i Or optical signal to noise ratio OSNR _i Is applied, P _i = C + D × ν _i Or OSNR _i = C + D × ν _i It becomes. When these relationships are expressed by a determinant, the following expression (1) is obtained.
[0051]
(Equation 1)

[0052]
Here, a weight matrix W as shown in the following equation (2), which indicates whether or not signal light of each optical frequency exists, for 80 WDM signal lights is introduced, and the least squares linear fitting method is applied. I do.
[0053]
(Equation 2)

[0054]
Where w in the above equation (2) _i For w when signal light is present _i = 1, w when not present _i = 0.
[0055]
Here, matrices A, b, and x shown in the following equation (3) are defined for the above equation (1).
[0056]
[Equation 3]

[0057]
When the weighted least squares equation is expressed using the relations of the above equations (1) to (3), the following equation (4) is obtained.
[0058]
(Equation 4)

[0059]
When this equation (4) is solved for x, the following equation (5) is obtained.
[0060]
(Equation 5)

[0061]
Therefore, the values of C and D can be obtained by calculating the determinant of equation (5).
[0062]
Each optical frequency ν monitored by the optical monitor unit 15 based on the relationship of the expression (5) by the least squares linear fitting method as described above. _i Optical power P corresponding to _i Or optical signal to noise ratio OSNR _i Is used to calculate the primary tilt component of the optical amplifier. Here, as an example of a specific calculation process, for the spectrum shown in FIG. 9, as shown in FIG. _i Optical power P corresponding to _i Are read, and the results of calculating the primary tilt component of the optical amplifier using the respective values are shown.
[0063]
First, when the values read from FIG. 10 are substituted into the above-described equation (1), the following equation (6) is obtained.
[0064]
(Equation 6)

[0065]
In addition, as shown in FIG. 9, since signal lights of all optical frequencies are present here, w ₁ ~ W ₈₀ Are all 1. From these relations, when x is obtained by calculating the above equation (5), the following equation (7) is obtained.
[0066]
(Equation 7)

[0067]
Therefore, with respect to the monitoring result shown in FIG. 9, the primary tilt component of the gain of the optical amplifier is a straight line having a primary slope value D = 0.893075 (dB / Hz) as shown by a solid line in FIG. Approximated.
[0068]
Next, the filter characteristics are optimized so that the first-order tilt component as described above is compensated for by the first compensator 11 '.
[0069]
Specifically, here, a case is considered in which, for example, a long-period filter having an FSR twice or more as large as the frequency band for performing gain equalization is used as the linear optical filter 11A 'used in the first compensator 11'. In this case, the optimization parameters of the long-period filter are the amplitude and phase of the filter characteristics. For this reason, it is necessary to obtain a relational expression between the amplitude and phase of the long-period filter and the inclination value D of the primary tilt component to optimize the filter characteristics. Such optimization of the linear optical filter 11A 'can be performed, for example, according to a calculation procedure to which a least squares linear fitting method as shown below is applied.
[0070]
The theoretical equation of the long-period filter used as the linear optical filter 11A 'is expressed by the following equation (8), assuming that it can be approximated by the theoretical equation of a pure periodic filter.
[0071]
(Equation 8)

[0072]
Here, s is the amplitude (true value) of the filter, Δν is the FSR (THz) of the filter, ν is the optical frequency, ν _b Is the bottom frequency (THz) of the filter. Note that the above bottom frequency is a value indicating the frequency when the attenuation of the filter becomes 0 dB.
[0073]
Also for the long-period filter described above, by applying a linear expression: y = C ′ + D ′ × x as in the case described above, F (ν _i ) = C ′ + D ′ × ν _i It becomes. With respect to this relational expression, the least squares linear fitting method is applied to 80 waves in which the signal light exists, and the determinant expressed by the following expression (9) is used.
[0074]
(Equation 9)

[0075]
With respect to the optimization of the filter characteristics of the first compensator 11 ′, F (ν _i ) Can be designed as if it exists, so there is no need to introduce a weight matrix W as shown in the above equation (2).
[0076]
Here, a matrix A, b ', x' shown in the following equation (10) is defined for the above equation (9).
[0077]
(Equation 10)

[0078]
When the least squares equation is expressed using the relations of the above equations (9) and (10), b ′ = Ax ′. When this least squares equation is solved for x ′, the following equation (11) is obtained.
[0079]
[Equation 11]

[0080]
By calculating the determinant of the equation (11), the values of C ′ and D ′ can be obtained.
[0081]
Further, as is apparent from the properties of the periodic filter, by controlling the phase of the filter, it is possible to realize two linear optical filters having the same absolute value of the first-order gradient from the filter having the required amplitude. Therefore, the absolute value | D | of the gradient of the primary tilt component of the optical amplifier calculated according to the above equation (5) and the absolute value | D '| of the slope of the periodic filter calculated according to the above equation (11) The filter amplitude s is optimized so that Actually, a relational expression between the amplitude s of the long-period filter and the absolute value D ′ of the primary gradient is obtained, and the amplitude s of the linear optical filter 11A ′ is set according to the relational expression so that | D | = | D ′ | It should be optimized.
[0082]
The relational expression between the amplitude of the long-period filter and the absolute value of the primary gradient is generally complicated. For this reason, when optimizing the filter amplitude s from the absolute value | D | of the primary gradient calculated according to the equation (5), it is preferable that the control unit 16 performs a search by, for example, the Newton method. In addition, regarding the phase of the filter, in order to minimize the average insertion loss of the filter, when the first-order gradient value D calculated according to the equation (5) is positive, the optical frequency ν ₁ And the bottom frequency ν of the filter _b Is equal to (ν ₁ = V _b On the other hand, when the primary slope value D is negative, the optical frequency ν ₈₀ And the bottom frequency ν of the filter _b Is equal to (ν ₈₀ = V _b ) Is desirably controlled.
[0083]
Here, a calculation process for optimizing the amplitude s of the linear optical filter 11A 'in accordance with the primary gradient value D = 0.893075 calculated in the above specific example will be described as an example.
[0084]
For example, the FSRΔν of the long-period filter used as the linear optical filter 11A ′ is 12 THz, and the bottom frequency ν _b Is 191.95 THz, the above equation (11) becomes the following equation (12).
[0085]
(Equation 12)

[0086]
From the above equation (12), in order to satisfy | D | = | D '|, it is necessary to satisfy the following equation (13).
[0087]
(Equation 13)

[0088]
If the above equation (13) is solved for s, the amplitude s of the linear optical filter 11A 'is obtained as 0.74974.
[0089]
FIG. 11 shows the filter characteristics of the linear optical filter 11A ′ optimized as described above. However, the vertical axis in FIG. 11 represents the attenuation in the linear optical filter 11A 'by a negative value. The WDM signal light that has passed through the linear optical filter 11A 'and compensated for the primary tilt component has a spectrum as shown in FIG. 12, for example. Here, also for the spectrum of FIG. _i Optical power P corresponding to _i Are read, and the primary tilt component is calculated according to the same procedure as above. As shown in FIG. 13, the slope D is substantially zero as shown in FIG. It can be seen that the compensation has been made. Further, as is apparent from a comparison between FIG. 10 and FIG. 13, it can be seen that the deviation of the signal light level at both ends of the signal light frequency band can be controlled to be extremely small.
[0090]
When the optimization of the filter characteristics of the first compensator 11 'is completed according to the above algorithm, the filter characteristics of the second compensator 12' are optimized.
[0091]
Specifically, the WDM signal light having passed through the first compensator 11 'and having the primary tilt component compensated passes through the second compensator 12' (the filter characteristics at this time are flat with respect to the optical frequency. State), a part of which is branched by the optical branching unit 14 as monitor light and sent to the optical monitoring unit 15. The optical monitor unit 15 detects the spectrum of the WDM signal light after the primary tilt component compensation as shown in FIG. _i Optical power P corresponding to _i And optical signal to noise ratio OSNR _i Is transmitted to the control unit 16.
[0092]
The control unit 16 performs Fourier series expansion on the optical frequency characteristic of the signal light power after the first-order tilt component compensation based on the monitoring result of the optical monitoring unit 15, and each of the plurality of periodic filters constituting the Fourier filter 12 A ′ Are optimized. As an algorithm for optimizing the second compensator 12 ′ based on the Fourier series expansion method, for example, a design method described in Japanese Patent Application Laid-Open No. 2000-199880 or the like can be applied. A brief description will be given.
[0093]
In the filter design using the Fourier series expansion method, when the FSR of the longest cycle filter among the plurality of periodic filters constituting the Fourier filter 12A ′ is 1, the FSR of the other periodic filters is 1 / n (where, n is an integer). At this time, the FSR of the filter having the longest period is designed by adding a margin frequency band to a signal light frequency band for performing gain equalization. In the prior art, the spare frequency band is usually set to about half of the signal light frequency band. However, in the gain equalizer 1b, since the deviation of the signal light level at both ends of the signal light frequency band is compensated by the first gain equalizer 11 ', the marginal frequency band can be made very narrow. it can. This makes it possible to reduce the number of stages of the Fourier filter 12A '.
[0094]
Specifically, considering the optical frequency characteristics of the signal light power after the first-order tilt component compensation as shown in FIG. 12 and FIG. 13, the signal light frequency band is 3.95 THz. It is possible to provide a little extra frequency band and set the FSR of the filter with the longest period to 5.1 THz. With this 5.1 THz as the basic period, the FSR of the other periodic filters is set to, for example, 2.55 THz, 1.7 THz, 1.275 THz, 1.02 THz, 0.85 THz and 0.729 THz, for a total of seven stages. The Fourier filter 12A 'is configured by a combination of the periodic filters. The amplitudes of the seven-stage periodic filters are individually set, for example, as shown in FIG. 14, according to the result of Fourier series expansion of the optical frequency characteristics of the signal light power. By combining the FSR and the seven-stage periodic filter whose amplitude is set in this way, the filter characteristic of the Fourier filter 12A 'is optimized to a shape corresponding to the sum of the seven-stage filters as shown in FIG. You. Here, the Fourier filter 12A 'having a seven-stage configuration is illustrated, but the number of stages of the Fourier filter in the present invention is not limited to this, and can be appropriately set according to the required compensation accuracy.
[0095]
The WDM signal light that has passed through the second compensator 12 ′ whose filter characteristics have been optimized as described above and whose undulation component has been compensated has, for example, a spectrum as shown in FIG. Further, regarding the spectrum of the WDM signal light shown in FIG. _i Optical power P corresponding to _i Are read and plotted, respectively, as shown in FIG. As is clear from FIGS. 16 and 17, the deviation between the optical frequencies of the signal light power is compensated with very high accuracy, and the WDM signal light has a flat signal light power over the entire signal light frequency band. Is obtained.
[0096]
After the optimization of each filter characteristic of the first compensator 11 'and the second compensator 12' at the time of startup as described above, the monitoring result of the optical monitor 15 is fed back to the controller 16 as needed. According to the change of the operation state of the optical amplifier or the change of the optical frequency used for the WDM signal light, the filter characteristics of the first compensator 11 ′ and the second compensator 12 ′ are changed according to the same algorithm as at the time of startup. Optimized.
[0097]
As described above, according to the gain equalizer 1b of the second embodiment, the same effect as that of the above-described first embodiment is obtained, and in addition, the first compensator 11 ′ and the second compensator 12 ′ are provided. By performing the feedback control by making each filter characteristic variable, it is possible to stably compensate the gain wavelength characteristic of the optical amplifier with higher accuracy.
[0098]
In the gain equalizers 1a and 1b of the first and second embodiments described above, the first compensators 11 and 11 'for compensating the primary tilt component of the gain wavelength characteristic of the optical amplifier are arranged on the input side, and the undulation component is provided. Are arranged on the output side to perform cascade connection. However, the arrangement of the first and second compensating sections is exchanged, and the second compensating sections 12, 12 ′ are arranged on the input side. It is also possible to arrange them and arrange the first compensating units 11 and 11 ′ on the output side for cascade connection.
[0099]
Next, an optical amplifier configured using the gain equalizers 1a and 1b of the first and second embodiments as described above will be described.
[0100]
FIG. 18 is a block diagram showing a configuration of the optical amplifier according to the third embodiment of the present invention.
[0101]
In FIG. 18, the present optical amplifier 2a is, for example, an optical amplifier having a known configuration in which the optical amplifiers 21 and 22 are connected in cascade to increase the optical output. In this case, the gain equalizer 1a of the first embodiment described above is applied.
[0102]
Specifically, the optical amplifier 2a includes an optical amplifier 21, a gain equalizer (GEQ) 1A, a variable optical attenuator (VOA) 23, and an optical amplifier 21, which are cascade-connected in sequence between an input terminal IN and an output terminal OUT. Amplifying section 22, optical branching section 24A and optical receiver (PD) 25A for monitoring input optical power to optical amplifying section 21, and optical branching section 24B for monitoring output optical power from optical amplifying section 21 And a light receiver 25B, an optical splitter 24C and a light receiver 25C for monitoring the input light power to the optical amplifier 22, and an optical splitter 24D and a light receiver for monitoring the output light power from the optical amplifier 22. And a control circuit 26 that controls the operation states of the optical amplifiers 21 and 22 and the variable optical attenuator 23 based on the monitoring results of the optical branching units 24A to 24D.
[0103]
The control circuit 26 controls the gain of each of the optical amplifiers 21 and 22 to be constant so that the gain wavelength characteristic of each of the optical amplifiers 21 and 22 does not change even when the input optical power to the optical amplifier 2a changes. AGC is performed, and the input dynamic range is controlled by controlling the amount of attenuation in the variable optical attenuator 23 so that the input power (and output power) to the optical amplifier 22 at the subsequent stage is constant. Note that the control method of the optical amplifier as described above is the same as the control method applied to the conventional optical amplifier, and the description is omitted here.
[0104]
In the optical amplifier 2a having the above-described configuration, the gain wavelength characteristics of the front and rear optical amplifiers 21 and 22 are compensated with high accuracy by the gain equalizer 1a according to the present invention inserted between the stages. become. Thereby, the present optical amplifier 2a can amplify the WDM signal light input to the input terminal IN to a desired level and output the WDM signal light having the same signal light level between the wavelengths from the output terminal OUT. .
[0105]
Next, an optical amplifier according to a fourth embodiment of the present invention will be described.
[0106]
FIG. 19 is a block diagram illustrating the configuration of the optical amplifier according to the fourth embodiment.
[0107]
In FIG. 19, the present optical amplifier 2b is, for example, provided with the gain equalizer 1b with feedback control described in the above-described second embodiment, for an optical amplifier having a known two-stage amplification configuration as described above. is there.
[0108]
Specifically, the optical amplifier 2b includes a first compensator 11 'and a second compensator 11' having a variable filter characteristic described in the above-described second embodiment between the optical amplifiers 21 and 22 at the front and rear stages. The compensating unit 12 'is cascaded, and a part of the amplified light output from the optical amplifying unit 22 at the subsequent stage is branched by the optical branching units 24D and 14 and sent to the optical monitoring unit 15. Based on the above, the filter characteristics of the first compensator 11 'and the second compensator 12' are feedback-controlled by the control circuit 26 '.
[0109]
In the above optical amplifier 2b, the gain wavelength characteristic of each of the optical amplifiers 21 and 22 can be dynamically compensated over a wide range by applying the gain equalizer 1b that performs the feedback control. It is not necessary to perform AGC on each of the optical amplifiers 21 and 22 as in the case of the embodiment, and it is sufficient to perform APC for controlling the output of each optical amplifier 21 and 22 to be constant. For this reason, the optical splitters 24A and 24C and the light receivers 25A and 25C which are arranged in front of the optical amplifiers 21 and 22 in the third embodiment can be omitted in the present embodiment. In addition, since it is not necessary to make the input power to the optical amplifier 22 in the subsequent stage constant, the variable optical attenuator 23 disposed between the optical amplifiers 21 and 22 in the third embodiment is also used in this embodiment. In the form, it can be omitted.
[0110]
As described above, according to the optical amplifier of the fourth embodiment, the gain equalizer 1b that performs feedback control of each filter characteristic of the first and second compensating units 11 'and 12' is applied, so that the configuration is stabilized with a simpler configuration. This makes it possible to perform the amplified light amplification.
[0111]
In the third and fourth embodiments described above, the configuration in which the first and second compensators are disposed between the optical amplifiers 21 and 22 in the preceding and subsequent stages has been described. However, the configuration of the optical amplifier of the present invention is not limited to this. It is not limited. For example, as in an optical amplifier 2c shown in FIG. 20, first and second compensators 11 'and 12' may be arranged between the input terminal IN and the optical amplifier 21 in the preceding stage. Further, for example, like the optical amplifier 2d shown in FIG. 21, the first and second compensating units 11 'and 12' can be arranged between the optical amplifying unit 22 and the optical branching unit 24D at the subsequent stage. However, in consideration of the level diagram, OSNR degradation, and the like, the insertion positions of the first and second compensating units 11 'and 12' are determined by the positions of the optical amplifying units 21 and 22 as in the optical amplifier 2b shown in FIG. It is desirable to be between.
[0112]
Next, a WDM optical transmission system configured using the above-described gain equalizers 1a and 1b or the optical amplifiers 2a to 2d will be described.
[0113]
FIG. 22 is a block diagram illustrating a configuration of a WDM optical transmission system according to a fifth embodiment of the present invention.
[0114]
The WDM optical transmission system 3a illustrated in FIG. 22 connects, for example, an optical transmission device 31 and an optical reception device 32 that transmit and receive WDM signal light with an optical transmission line 33 and has a predetermined interval on the optical transmission line 33. In the known system configuration in which the optical amplifier 34 is arranged and the WDM signal light is relayed and transmitted, the gain equalizer 1a or the gain equalizer 1a or the 1b (shown as GEQ1 in the figure).
[0115]
In such a WDM optical transmission system 3a, the gain wavelength characteristic of the optical amplifier 34 in each relay section is compensated with high accuracy by the corresponding gain equalizer (GEQ) 1, so that the optical transmitter 31 In addition, the relay transmission of the WDM signal light between the optical receivers 32 can be performed reliably and stably.
[0116]
In the WDM optical transmission system of the fifth embodiment, the gain equalizer of the present invention is provided in correspondence with the general optical amplifier 34. However, the optical amplifier of each relay section in the known system configuration is used. As 34, for example, any one of the optical amplifiers 2a to 2d (shown as the optical amplifier 2 in the figure) shown in the third and fourth embodiments as in the WDM optical transmission system 3b shown in FIG. You may.
[0117]
Also, an example has been shown in which a gain equalizer is arranged for each relay section to perform gain equalization. However, gain equalization for two or more relay sections is collectively performed by one gain equalizer. Is also possible. In this case, for example, as in the WDM optical transmission system 3c shown in FIG. 24, the first compensator 11 'and the second compensator 12' are arranged in different relay sections, that is, the first compensator 11 'and the second compensator 11' The arrangement is such that the optical amplifier 34 and the optical transmission line 33 exist between the sections 12 ′, and the filter characteristics of the first and second compensating sections 11 ′ and 12 ′ in different relay sections are fed back by the control section 16 in conjunction with each other. It is also possible to control.
[0118]
The main inventions disclosed in the present specification are summarized below.
[0119]
(Supplementary Note 1) A gain equalizer that compensates for wavelength characteristics of light,
A first compensator that has a loss wavelength characteristic that monotonically increases or monotonically decreases with respect to the compensation target wavelength region, and compensates for a linear gain component of the wavelength characteristic in the compensation target wavelength region;
A second compensator for compensating a wavelength characteristic in the compensation target wavelength region that is not compensated by the first compensator by a combination of a plurality of periodic filters cascaded with the first compensator;
A gain equalizer characterized by comprising:
[0120]
(Supplementary Note 2) The gain equalizer according to Supplementary Note 1, wherein
The first compensator and the second compensator each have variable loss wavelength characteristics,
An optical monitoring unit that monitors light passing through the first compensation unit and the second compensation unit;
A control for judging a wavelength characteristic in the compensation target wavelength region based on a monitoring result of the optical monitoring unit, and performing feedback control of each loss wavelength characteristic of the first compensating unit and the second compensating unit in accordance with the judgment result. Department and
A gain equalizer characterized by comprising:
[0121]
(Supplementary Note 3) The gain equalizer according to Supplementary Note 1, wherein
The first compensator has a long-spectrum filter having a free spectrum region wider than the wavelength region to be compensated, and a periodic loss wavelength characteristic monotonically increasing or decreasing in the wavelength region to be compensated. Gain equalizer.
[0122]
(Supplementary Note 4) The gain equalizer according to Supplementary Note 1, wherein
The gain equalizer, wherein the second compensator includes an optical filter configured by combining a plurality of periodic filters having different free spectral regions.
[0123]
(Supplementary note 5) The gain equalizer according to supplementary note 2, wherein
The control unit calculates a first-order slope value of a wavelength characteristic in the compensation target wavelength region by applying a least-squares linear fitting method to a monitoring result of the optical monitoring unit, and the loss wavelength characteristic of the first compensation unit is the A gain equalizer, wherein feedback control is performed on the first compensator so as to monotonically increase or decrease at a rate corresponding to the primary slope value.
[0124]
(Supplementary Note 6) The gain equalizer according to Supplementary Note 2, wherein
The control unit applies a Fourier series expansion method to a monitoring result of the optical monitor unit, and controls the second compensating unit so that an amplitude and a phase of each periodic filter of the second compensating unit correspond to a calculation result of the Fourier series expansion. A gain equalizer performing feedback control for a compensator.
[0125]
(Supplementary note 7) The gain equalizer according to supplementary note 2, wherein
The optical monitoring unit monitors signal light power corresponding to each wavelength of the wavelength multiplexed signal light,
The gain equalizer, wherein the control unit determines a wavelength characteristic in the compensation target wavelength region by arithmetic processing using signal light power of each wavelength monitored by the optical monitor unit.
[0126]
(Supplementary note 8) The gain equalizer according to supplementary note 2, wherein
The optical monitoring unit monitors an optical signal-to-noise ratio corresponding to each wavelength of the wavelength multiplexed signal light,
The gain equalizer, wherein the control unit determines a wavelength characteristic in the wavelength region to be compensated by an arithmetic process using an optical signal-to-noise ratio of each wavelength monitored by the optical monitor unit.
[0127]
(Supplementary Note 9) An optical amplification unit that collectively amplifies the wavelength multiplexed signal light,
A first compensator that has a loss wavelength characteristic that monotonically increases or monotonically decreases with respect to the wavelength region to be compensated, and compensates for a linear gain component of the gain wavelength characteristic of the optical amplifier;
A second compensator for compensating for a gain wavelength characteristic of the optical amplifier, which is not compensated for by the first compensator, by a combination of a plurality of periodic filters cascaded to the first compensator;
An optical amplifier, comprising:
[0128]
(Supplementary note 10) The optical amplifier according to supplementary note 9, wherein
The first compensator and the second compensator each have variable loss wavelength characteristics,
An optical monitoring unit that monitors the wavelength multiplexed signal light that has passed through the first compensation unit and the second compensation unit;
A control unit that determines a gain wavelength characteristic of the optical amplifying unit based on a monitoring result of the optical monitoring unit and performs feedback control of each loss wavelength characteristic of the first compensating unit and the second compensating unit according to the determined result When,
An optical amplifier, comprising:
[0129]
(Supplementary Note 11) The optical amplifier according to supplementary note 9, wherein
The optical amplifier, wherein the first compensator and the second compensator compensate each gain wavelength characteristic of a plurality of cascade-connected optical amplifiers.
[0130]
(Supplementary Note 12) The optical amplifier according to supplementary note 11, wherein
The optical amplifier according to claim 1, wherein the first compensator and the second compensator are disposed between the cascade-connected optical amplifiers.
[0131]
(Supplementary Note 13) An optical transmission system that amplifies wavelength multiplexed signal light collectively by an optical amplifier disposed on an optical transmission line, and compensates for a gain wavelength characteristic of the optical amplifier by a gain equalizer, and performs relay transmission. hand,
The gain equalizer has a loss wavelength characteristic that monotonically increases or decreases in a compensation target wavelength region, and a first compensator that compensates for a linear gain component of the gain wavelength characteristic of the optical amplifier; An optical transmission system comprising: a second compensator for compensating a gain wavelength characteristic of the optical amplifier that is not compensated for by the first compensator by a combination of a plurality of periodic filters cascaded to the compensator. .
[0132]
(Supplementary note 14) The optical transmission system according to supplementary note 13, wherein
In the gain equalizer, the first compensator and the second compensator each have a variable loss wavelength characteristic,
An optical monitoring unit that monitors the wavelength multiplexed signal light that has passed through the first compensation unit and the second compensation unit;
A control unit for determining a gain wavelength characteristic of the optical amplifier based on a monitoring result of the optical monitoring unit, and performing feedback control of each loss wavelength characteristic of the first compensation unit and the second compensation unit according to the determination result; ,
An optical transmission system, comprising:
[0133]
(Supplementary note 15) The optical transmission system according to supplementary note 14, wherein
When a plurality of optical amplifiers are arranged on the optical transmission line and the wavelength multiplexed signal light is transmitted through a plurality of relay sections,
The optical transmission system according to claim 1, wherein the gain equalizer is disposed in each relay section.
[0134]
(Supplementary note 16) The optical transmission system according to supplementary note 14, wherein
When a plurality of optical amplifiers are arranged on the optical transmission line and the wavelength multiplexed signal light is transmitted through a plurality of relay sections,
The optical transmission system of the gain equalizer, wherein the first compensator and the second compensator are arranged in different relay sections.
[0135]
(Supplementary Note 17) A gain equalization method for compensating for wavelength characteristics of light,
The first compensator having a monotonically increasing or monotonically decreasing loss wavelength characteristic for the compensation target wavelength region compensates for a linear gain component of the wavelength characteristic in the compensation target wavelength region,
A second compensator comprising a combination of a plurality of periodic filters connected in cascade to the first compensator compensates for wavelength characteristics in the compensation target wavelength region which are not compensated by the first compensator. Gain equalization method.
[0136]
(Supplementary note 18) The gain equalization method according to supplementary note 17, wherein
Monitoring the light passing through the first compensator and the second compensator,
Based on the monitoring result, determine a wavelength characteristic in the compensation target wavelength region,
A gain equalization method, wherein feedback control is performed on each loss wavelength characteristic of the first compensator and the second compensator in accordance with the result of the determination.
[0137]
【The invention's effect】
As described above, according to the gain equalization technique of the present invention, for the light to be subjected to gain equalization, the linear gain component of the wavelength characteristic in the compensation target wavelength region is compensated by the first compensator, and the Since the swell component that has not been compensated by the first compensator is compensated by the second compensator, gain equalization can be efficiently performed with high accuracy.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of a gain equalizer according to a first embodiment of the present invention.
FIG. 2 is a diagram showing an example of a specific configuration of a linear optical filter used in the first embodiment.
FIG. 3 is a diagram illustrating an example of a gain wavelength characteristic of an optical amplifier to be subjected to gain equalization in the first embodiment.
FIG. 4 is a diagram showing gain wavelength characteristics after a first-order tilt component has been compensated by a first compensator in the first embodiment.
FIG. 5 is a diagram showing a loss wavelength characteristic of a Fourier filter designed according to the gain wavelength characteristic of FIG. 4;
6 is a diagram showing a loss wavelength characteristic of a Fourier filter (prior art) designed according to the gain wavelength characteristic of FIG. 3;
FIG. 7 is a block diagram illustrating a configuration of a gain equalizer according to a second embodiment of the present invention.
FIG. 8 is a diagram showing an example of a specific configuration of a linear optical filter used in the second embodiment.
FIG. 9 is a diagram showing an example of a spectrum of a WDM signal light monitored at the time of startup in the second embodiment.
FIG. 10 is a diagram showing a result of reading optical power corresponding to an optical frequency and calculating a primary tilt component in FIG. 9;
11 is a diagram illustrating filter characteristics of a linear optical filter optimized according to the result of FIG. 10;
12 is a diagram illustrating a spectrum of a WDM signal light that has passed through the linear optical filter of FIG. 11 and has been compensated for a primary tilt component.
FIG. 13 is a diagram showing a result of reading optical power corresponding to an optical frequency and calculating a primary tilt component in FIG. 12;
FIG. 14 is a diagram illustrating a characteristic example of a plurality of periodic filters forming a Fourier filter in the second embodiment.
FIG. 15 is a diagram illustrating characteristics of a Fourier filter obtained by combining the periodic filters in FIG. 14;
FIG. 16 is a diagram illustrating a spectrum of a WDM signal light that has passed through the Fourier filter of FIG. 15 and has a swell component compensated.
17 is a diagram showing a result of reading optical power corresponding to an optical frequency in FIG.
FIG. 18 is a block diagram showing a configuration of an optical amplifier according to a third embodiment of the present invention.
FIG. 19 is a block diagram showing a configuration of an optical amplifier according to a fourth embodiment of the present invention.
FIG. 20 is a block diagram showing another configuration example related to the fourth embodiment.
FIG. 21 is a block diagram showing another configuration example related to the fourth embodiment.
FIG. 22 is a block diagram illustrating a configuration of a WDM optical transmission system according to a fifth embodiment of the present invention.
FIG. 23 is a block diagram showing another system configuration example related to the fifth embodiment.
FIG. 24 is a block diagram showing another example of the system configuration related to the fifth embodiment.
FIG. 25 is a diagram for explaining a problem with a conventional gain equalizer.
FIG. 26 is a diagram for explaining another problem related to the conventional gain equalizer.
[Explanation of symbols]
1A, 1B gain equalizer (GEQ)
11, 11 'First compensation unit
11A Linear optical filter
12, 12 'Second compensator
12A Fourier filter
13 Light Path
14 Optical branching section
15 Optical monitor
16 control unit
2A-2D optical amplifier
21,22 Optical amplifier
23 Variable Optical Attenuator (VOA)
24A to 24D Optical branching unit
25A to 25D photo detector (PD)
26 Control circuit
IN input terminal
OUT output terminal

Claims (5)

A gain equalizer that compensates for wavelength characteristics of light,
A first compensator that has a loss wavelength characteristic that monotonically increases or monotonically decreases with respect to the compensation target wavelength region, and compensates for a linear gain component of the wavelength characteristic in the compensation target wavelength region;
A second compensator for compensating a wavelength characteristic in the compensation target wavelength region that is not compensated by the first compensator by a combination of a plurality of periodic filters cascaded with the first compensator;
A gain equalizer characterized by comprising: The gain equalizer according to claim 1, wherein
The first compensator and the second compensator each have variable loss wavelength characteristics,
An optical monitoring unit that monitors light passing through the first compensation unit and the second compensation unit;
A control for judging a wavelength characteristic in the compensation target wavelength region based on a monitoring result of the optical monitoring unit, and performing feedback control of each loss wavelength characteristic of the first compensating unit and the second compensating unit in accordance with the judgment result. Department and
A gain equalizer characterized by comprising: The gain equalizer according to claim 2, wherein
The control unit calculates a first-order slope value of a wavelength characteristic in the compensation target wavelength region by applying a least-squares linear fitting method to a monitoring result of the optical monitoring unit, and the loss wavelength characteristic of the first compensation unit is the A gain equalizer, wherein feedback control is performed on the first compensator so as to monotonically increase or decrease at a rate corresponding to the primary slope value. An optical amplifying unit that amplifies the wavelength multiplexed signal light collectively;
A first compensator that has a loss wavelength characteristic that monotonically increases or monotonically decreases with respect to the wavelength region to be compensated, and compensates for a linear gain component of the gain wavelength characteristic of the optical amplifier;
A second compensator for compensating for a gain wavelength characteristic of the optical amplifier, which is not compensated for by the first compensator, by a combination of a plurality of periodic filters cascaded to the first compensator;
An optical amplifier, comprising: An optical transmission system for amplifying wavelength-multiplexed signal light collectively by an optical amplifier disposed on an optical transmission path, and for relay transmission by compensating a gain wavelength characteristic of the optical amplifier by a gain equalizer,
The gain equalizer has a loss wavelength characteristic that monotonically increases or decreases in a compensation target wavelength region, and a first compensator that compensates for a linear gain component of the gain wavelength characteristic of the optical amplifier; An optical transmission system comprising: a second compensator for compensating a gain wavelength characteristic of the optical amplifier that is not compensated for by the first compensator by a combination of a plurality of periodic filters cascaded to the compensator. .

Images