In recent years, the digital coherent receiver have been studied extensively in the field of optical fiber transmission. Since such a receiver can restore the full information of the optical electric field, we can realize spectrally-efficient systems using polarization-division multiplexing and multilevel-modulation formats such as quadrature-phase-shift keying (QPSK) and higher-order quadrature-amplitude modulation (QAM). On the other hand, system impairments have to be overcome in order to elongate the maximum transmission distance of such signals. The most crucial system impairments are the laser phase noise, the frequency offset, chromatic dispersion, polarization-mode dispersion, polarization-dependent loss/gain, and fiber nonlinearity. The digital coherent optical receiver enables post-compensation for these impairments with sophisticated digital-signal processing (DSP). Such post-compensation is one of the most important functions of the digital coherent optical receiver, and this dissertation is dedicated to the research and development of post-compensation schemes and to realize ultra-long haul transmission systems with high spectral efficiency.

Recently, the digital coherent optical receiver for the dual-polarization QPSK signal has been introduced to the commercial system. However, most of the demodulation algorithms employed in the QPSK receiver cannot directly be applied to the higher-order QAM signals. In order to achieve further spectrally-efficient systems, novel demodulation algorithms for higher-order QAM signals are indispensable. Another issue is that fiber-nonlinearity compensation is not conducted in such receivers since it requires the high computational cost. To achieve long-haul transmission distance with higher-order QAM signals, we must compensate for fiber nonlinearity.

Considering these problems inherent in the current digital coherent receivers, we propose novel demodulation algorithms applicable to all of transmission systems in a unite manner. The results of the research are as follows.

First, we propose a novel dual-stage decision-directed phase estimator, which is based on the one-tap finite-impulse-response (FIR) filters and the decision-directed least-mean-square (DD-LMS) algorithm with the training mode. This scheme can eliminate phase fluctuations induced both by the phase noise and by the frequency offset in all of multilevel-modulation systems; in particular, it enables perfect frequency-offset elimination without any power penalty as far as the whole signal spectrum is properly converted into the digital domain. Our proposed scheme consists of the first-stage estimator tracking the phase noise and the second-stage estimator tracking the frequency offset. These estimators are designed so as to operate without mutual interaction with the following procedures: (1) The fast phase change by the frequency offset is estimated from the second-stage estimator and is removed from the error signal in the DD-LMS algorithm controlling the first-stage estimator. (2) The frequency offset estimated from the second-stage estimator is eliminated from the tap coefficient of the first-stage estimator by using a phase integrator. Using such a scheme, the proposed phase estimator can work in a stable manner even with frequency offsets much larger than the symbol rate. Based on computer simulations and experiments, the effectiveness of the proposed phase estimator is confirmed in higher-order QAM systems.

Second, we propose a novel configuration of the FIR filter adapted by the DD-LMS algorithm with the training mode, in which fast phase fluctuations are removed from the error signal used for tap adaptation of FIR filters. With such a scheme, we can employ arbitrary-long delay taps in FIR filters even under the influence of fast phase fluctuations. Our proposed scheme can operate under polarization-mode dispersion and polarization-dependent loss/gain without any singularity problem owing to the training sequence for initialization of the tap coefficients. In addition, our scheme can easily be applied to any multilevel modulation formats, since it employs the DD-LMS algorithm in all of the tap-adaptation processes. Through computer simulations and experiments, we show that the proposed FIR-filtering scheme is robust against the phase noise and the frequency offset in higher-order QAM systems.

Third, we propose a novel Kerr-effect compensator with the low computational cost. In order to improve the performance of a split step, we modify a configuration of a single split-step compensator by employing dispersive elements and filters. Using such a modified split step, we can achieve the best demodulation performance with the lowest computational cost. Next, we propose a novel Kerr-effect-compensation scheme, in which the modified split steps are aligned in parallel in the frequency domain. With such a scheme, parallel signal processing is achieved, which is suitable for hardware implementation. We confirm that the proposed scheme significantly elongates the maximum transmission distance of QAM signals both in dispersion-managed systems and in dispersion-unmanaged systems.

Our proposed post-compensation schemes can be applied to any transmission systems. The results we obtained open up the possibility of ultra-long haul transmission systems with high capacity and have enormous significance in the development of future optical fiber networks.

本論文は,Digital Compensation for Transmission Impairments in Coherent Optical Receivers (コヒーレント光受信器におけるディジタル信号処理を用いた伝送制限要因の補償に関する研究)と題し,英文で執筆されており,7章からなる。コヒーレント光受信器は光電界の全ての情報を保持できるので,受信信号のディジタル信号処理により,送受信および伝送によって生じる信号劣化の補償が可能となる。本研究では,多値光変調信号の超長距離伝送を目的として,あらゆる多値変調方式に対応できる復調アルゴリズムの提案を行い,計算機シミュレーションおよび実験により,これらのアルゴリズムの有効性を検証している。

第1章はIntroductionであり,ディジタルコヒーレント光受信器の特長を論じたのち,本論文の目的と構成について述べる。

第2章はCoherent optical receiversと題し,コヒーレント光検出の原理と受信器の構成についてまとめている。

第3章はDigital compensation for system impairmentsと題し,コヒーレント光受信器に導入されているディジタル信号処理アルゴリズムの概要について述べている。

第4章はDual-stage decision-directed phase estimator enabling perfect frequency-offset eliminationと題し,零次の有限インパルス応答(FIR)フィルタと判定指向型最小二乗平均(DD-LMS)アルゴリズムに基づく新たな位相推定器を提案している。本位相推定器は,位相雑音と周波数オフセットによる位相変動を分離して推定することにより,高い周波数オフセット耐力が実現できる。本手法の有効性は,計算機シミュレーションおよび実験によって確認された。

第5章はPhase-fluctuation-tolerant FIR-filter configuration based on DD-LMS algorithmと題し,線形歪み補償のための,DD-LMSアルゴリズムに基づく新たなFIRフィルタ構成を提案した。本構成を用いることにより,低速な線形歪みを除去する線形等化器と高速な位相変動を除去する位相推定器の適応操作を分離し,それぞれを安定に動作させることが可能となる。また本手法は,全ての適応過程にDD-LMSアルゴリズムを使用しているため,あらゆる多値変調方式に容易に適用できるという特長を持つ。計算機シミュレーションおよび実験によって,位相雑音および周波数オフセットが存在しても,提案手法は安定に動作することを確認した。

第6章はKerr-effect compensator with parallel split stepsと題し,ハードウェア実装に適した新たなカー効果補償器を提案した。まず,計算コストを減少させるため,単一スプリットステップで構成される補償器の性能改善を行った。さらに,この改善したスプリットステップを周波数領域で並列展開することにより,高シンボルレートの信号に生じたカー効果歪みを,並列処理によって除去する補償方式を提案した。計算機シミュレーションによって,本手法は分散管理光ファイバ伝送路における多値光変調信号の伝送可能距離を飛躍的に延長できることを確認した。

第7章はConclusionであり,本論文で得られた成果をまとめ,今後の展望について述べている。

以上のように本研究では,ディジタルコヒーレント光受信器への導入を目的として,新たな復調アルゴリズムに基づき,高い周波数オフセット耐力を持つ位相推定器,高速な位相変動のもとでも安定に動作する線形歪み補償用FIRフィルタ,ハードウェア実装に適したカー効果補償器を提案し,その有効性を計算機シミュレーションおよび実験により示した。これらの成果は,多値光変調信号の超長距離伝送を可能とし,将来の大容量光ネットワークの発展に寄与するもので,電子工学への貢献が多大である。

よって本論文は,博士(工学)の学位請求論文として合格と認められる。