基于单个半导体光放大器的高速多功能逻辑门

采用半导体光放大器(SOA)中的非线性效应可以实现多种多样的高速全光信号处理. 利用SOA的非线性效应(包括四波混频、交叉增益调制、瞬态交叉相位调制等)实现了多种功能的逻辑运算,包括“与”、“或非”、“同或”、“或”和“非”. 由于SOA用于全光信号处理的调制速率受到增益恢复时间较慢的限制而无法实现高速的信号处理,在SOA后面级联一个带宽为0.32nm的失谐滤波器可以提高SOA的工作速率,仅用一个SOA实现了40Gbit/s的多功能逻辑门. 关键词： 半导体光放大器 全光逻辑门 瞬态交叉相位调制     

All optical wavelength converters based on cross phase modulation in SOA-MZI configuration

Wavelength converter plays an important role for increasing the capacity and flexibility of future broadcast network. The cross phase modulation (XPM) based converter has high conversion efficiency at low input power. In order to improve the efficiency and wideband conversion range, the XPM is increase by optimizing the semiconductor optical amplifiers-Mach–Zehnder interferometer (SOA-MZI) configuration. The XPM is improved by increasing the active region length and bias current of the SOA.     

基于UNI开关的40 GHz注入锁模全光时钟提取实验

基于超快非线性干涉仪UNI开关具有超过40 Gbit/s的开关速度，提出一种在锁模环形腔内利用UNI开关进行时钟提取的新方案.按此方案进行实验，对40 Gbit/s伪随机信号进行时钟提取，得到了清晰稳定的时钟脉冲.与利用SOA的交叉增益调制注入锁模时钟提取方案相比，在此方案中注入到SOA的信号功率仅0.5 dBm，降低了7 dB左右，对SOA的运行更安全.     

基于半导体光放大器和光学滤波器的同相波长转换理论研究

采用半导体光放大器(SOA)级联滤波器的方案可以实现高速全光波长转换,其中,滤波器实质上是对SOA输出信号的光谱进行优化操作.文中利用带通滤波器(BPF)和延时干涉仪(DI)组成的光谱优化器,不仅实现了80Gbit/s同相波长转换,而且基于遗传算法设计光谱优化器的参数,极大地提高了输出信号的质量.同时,也从理论上分析了SOA中超快带内效应对信号光谱的影响以及光谱优化器中BPF和DI的消光比对输出信号质量的影响.     

一种全光归零码到非归零码变换的新技术方案

报道了一种利用单只半导体光放大器和光滤波器实现全光归零码到非归零码变换的新技术方案.当探测（Probe）光和数据信号光同时输入到SOA时,基于SOA中的交叉增益调制和交叉相位调制现象,探测光光谱的后沿和前沿将分别产生红移和蓝移.通过调节光滤波器和探测光的中心波长之间的失调量,滤出光谱的特定部分,可以得到转换后的NRZ码光信号.这种新型的全光码型变换器具有结构简单、偏振不敏感、控制参量少和稳定性高的特点.分别采用仿真和实验的方式实现了20 Gbit/s光数据信号从RZ码到NRZ码的全光码型变换,并且仿真结果和实验结果相吻合.     

Photonic generation of millimeter-wave ultra-wideband signal using phase modulation to intensity modulation conversion and frequency up-conversion

We propose and experimentally demonstrate the generation of a pair of polarity-reversed 24 GHz millimeter-wave (MMW) ultra-wideband (UWB) monocycles. The scheme is realized by using delay interferometer (DI) based phase modulation to intensity modulation (PM–IM) conversion and carrier suppression modulation (CSM) based frequency up-conversion. The phase modulation is realized by using either electro-optic phase modulator (EOPM) or cross phase modulation (XPM) in semiconductor optical amplifier (SOA), which is an all-optical approach to obtaining baseband UWB signals, respectively. After frequency up-converted by using CSM in a Mach–Zehnder modulator (MZM), a pair of polarity-reversed 24 GHz MMW-UWB signals complying with the Federal Communication Committee (FCC) requirements is generated. The bi-phase modulation (BPM) of 24 GHz MMW-UWB signals can also be realized by electrically switching the bias voltage of delay interferometer.     

Characterization and optimization of SOA-based UNI recirculating shift register switching window

The operation of an all-optical recirculating shift register implemented with the semiconductor optical amplifier (SOA)-based ultrafast nonlinear interferometer (UNI) in a counter-propagating, inverting configuration is methodically studied and analyzed. This is achieved by applying an appropriate model for a SOA deployed as the nonlinear element in an interferometric switch driven by ultrafast and strong feedback optical pulses. By means of numerical simulation the impact of the critical functional parameters on the formation of the switching window is thoroughly investigated and assessed enabling to provide useful design rules for their proper selection and combination so as to optimize the specific metric and ensure high performance. The undertaken theoretical treatment can be extended for characterizing other more complex all-optical circuits and subsystems of enhanced functionality that employ this type of shift register as the key building module.     

All-optical signal regeneration using fiber nonlinearity

All-optical signal processing requires the use of nonlinear optical materials. One promising candidate for this purpose for operation at speeds faster than a few tens of Gbit/s is the optical fiber. In this paper, we describe all-optical signal regeneration using the self-phase modulation (SPM) in nonlinear fibers. Our recent experimental work on high-speed on-off keying signal regeneration using a two-stage SPM-based regenerator in bidirectional fiber configuration and on its application to differential phase-shift-keying signal regeneration is described in detail.     

Ultrafast all-optical signal processing with Symmetric Mach–Zehnder type all-optical switches

We present Symmetric Mach–Zehnder (SMZ) type all-optical switches: an SMZ all-optical switch, a polarization-discriminating SMZ (PD-SMZ) all-optical switch, and a delayed interference signal wavelength converter (DISC). These switches are capable of ultrafast, low control power, and low chirp switching, which is not restricted by slow relaxation of highly efficient nonlinearities. High repetition operation unrestricted by slow relaxation is also possible for these switches. This is because of a push–pull modulation scheme or sometimes called a differential phase modulation scheme. These three devices are similar, but different in some important aspects, thus a comparison is made among the three. Then semiconductor optical amplifiers (SOAs) are discussed as a nonlinear phase shifter for these devices. Then, ultrafast all-optical signal processing using SOA based SMZs is demonstrated. Error-free demultiplexing from 168 to 10.5 Gbit/s is presented, in which a hybrid-integrated SMZ (HI-SMZ) is used as a demultiplexer. In pulse regeneration experiment, the signal pulses at 84 Gbit/s are regenerated by the PD-SMZ and the regenerated pulses are demultiplexed to 10.5 Gbit/s by the HI-SMZ to verify error-free operation. The retiming capability of this scheme is quantitatively demonstrated. Also presented is error-free all-optical wavelength conversion at 168 Gbit/s using the DISC. These results represent the fastest error-free operations reported to date in each category.     

基于折叠式超快非线性干涉仪的全光码型转换理论和实验研究

码型转换是实现非归零(NRZ)信号全光时钟恢复的关键技术.提出了一种基于折叠式超快非线性干涉仪的全光NRZ到伪归零(PRZ)信号转换的新型方案.基于半导体光放大器的分段模型和超快非线性干涉仪的传输函数，建立了该方案的理论分析模型，数值模拟了不同速率下的码型转换过程.实验实现了稳定的10，20和40 Gbit/s的NRZ到PRZ信号的码型转换，分析了影响码型转换输出特性的因素，理论分析结果与实验结果基本符合. 关键词： 超快非线性干涉仪 半导体光放大器 全光码型转换     

半导体光放大器的光-光互作用及其应用

半导体光放大器（SOA）中的非线性系数约为普通光纤的10^9，为光子晶体光纤的10^7，而且有4种光-光互作用，即交叉增益调制（XGM）、交叉相位调制（XPM）、交叉偏振调制（XSM）及四波混频（FWM），可以灵活地组成各种光信号处理器件，如波长变换器、全光触发器、全光逻辑、全光时钟恢复、全光缓存器……等，正成为整个光信号处理的基础。文章介绍了它们的原理和简单应用。     

Q-factor assessment of SOA-based ultrafast nonlinear interferometer

The performance of the semiconductor optical amplifier (SOA)-based ultrafast nonlinear interferometer (UNI) gated by an ultra high speed pseudorandom binary sequence is theoretically analyzed. For this purpose, a comprehensive model formed by a set of equations that describe the gain and phase evolution inside a SOA deployed as the nonlinear element in an interferometric switch is appropriately applied to this particular configuration. By undertaking a detailed numerical simulation, the impact of the SOA and input data key parameters on the Q-factor is thoroughly investigated and assessed enabling to extract useful design rules for their proper selection so as to optimize this metric. The calculations confirm the experimental evidence that the main technical limitation is imposed by the SOA carrier lifetime, which must be reduced below the bit period in order to avoid the deleterious consequences of the pattern effect on the switched-out pulses. Provided that this condition is satisfied and the rest of the parameters fulfill their specified requirements, the output amplitude fluctuations can be effectively eliminated resulting in a high quality eye diagram and error-free operation. The adopted model can be exploited for studying more sophisticated all-optical circuits and subsystems of enhanced functionality that rely critically on the SOA-based UNI as switching module.     

基于光纤光参量放大的异步双波长全光再生技术研究

本文提出了一种新型的多波长全光再生方案,利用相位时钟光纤光参量放大,并采用相邻信道偏振正交的方法,实现对由异步信源产生的双波长信号全光再生.理论分析了参量放大中的增益饱和现象用于幅度噪声抑制,以及利用相位时钟及后续色散实现对信号定时的机理.在这个基础上,对两个独立信源产生的异步双波长10Gbit/s信号进行再生实验,实验表明该方案有效的抑制了基于多波长3R再生系统中信道间的四波混频与交叉相位调制等非线性干扰.系统在单波长和双波长情况下分别将两路信号信噪比改善了至少6.5dB与4.5dB.误码率测试结果说明,与背对背测试结果相比,无论是在单波长还是双波长条件下,两路波长的信号经过再生后都实现了约2dB的接收机功率代价的改善.     

基于半导体光放大器的环路型全光缓存器中功率均衡的新方法

全光缓存器能够在光域内对数据包进行存储而不需经过光电光的变换,成为全光网的重要组成部分,其性能的优劣将直接影响到网络的丢包率等。但在半导体光放大器中交叉增益调制和交叉相位调制效应并存,导致基于半导体光放大器的环路型全光缓存器中缓存后输出的数据包与未缓存数据包的输出功率间存在不均衡,使得网络误码率增加。在详细分析全光缓存器工作原理的基础上,利用半导体光放大器中交叉增益调制和交叉相位调制并存的现象,巧妙地提出了将控制脉冲反相并配合电可调衰减器的新方法。该方法简单易行,有效解决了此类全光缓存器中数据包输出功率不均衡的难题,缓存器性能得到了极大改善。此方法适用于所有基于半导体光放大器的环路型全光缓存器。     

SOA-assisted Sagnac switch and investigation of its roadmap from 10 to 40 GHz

Future broadband optical communications networks will rely on all-optical switches to perform a set of processing functions exclusively in the optical domain. Interferometric optical switches using semiconductor optical amplifier (SOA) nonlinearities can perform efficient optical switching with few tens of fJ control energies and short fiber lengths allowing for monolithic integration. In this paper we present work performed with a three terminal SOA-assisted Sagnac interferometer. We review all-optical Boolean AND and XOR logic results at 10 GHz and 10 Gb/s for full duty cycle and pseudo-random data pattern operation, respectively, achieved with adequate contrast ratios, remarkably low switching energies and low pattern dependence. The ability of the device to be cascaded was proved up to 10 GHz by recirculating stably for hours arbitrary pattern profiles. Finally, and in view of the extension of photonic networks single channel data rates beyond 40 Gb/s, the performance of the switch was simulated in terms of its critical parameters. The obtained results showed that full switching operation at 40 GHz or higher is feasible either by deploying gain recovery reduction techniques in bulk and quantum well SOAs or alternatively other technologically advanced optical devices, such as quantum dot SOAs.     

基于半导体光放大器和光滤波器的全光非归零码到归零码的转换

利用半导体光放大器模型和仿真软件对全光非归零码到归零码的变换进行了数值仿真.在仿真结果的基础上,实现了基于半导体光放大器和光滤波器的10 Gbps的全光非归零码到归零码的变换试验.试验结果显示在RZ码输入功率为－15 dBm时,该变换的误码率为1.0×10－9.     

Simultaneous all-optical digital comparator and dual-directional half-subtractor for two-input 40 Gbit/s DPSK signals employing SOAs

A module of simultaneous implementation of all-optical digital comparator and dual-directional half-subtractor is proposed. Proof-of-concept experiment is performed at 40 Gbit/s employing the four-wave mixing and cross gain modulation in three parallel semiconductor optical amplifiers. All output results with over 10 dB extinction ratios, clear and wide open eye diagrams, are obtained without using assistant/holding light beam. All-optical half-adder can also be obtained by adjusting the phase shifter of delay interferometer in the proposed module because of its inherent reconfigurability and flexibility. The module would be a promising digital logic elementary circuit in all-optical networks and computing systems.     

Photonic multi-shape UWB pulse generation using a semiconductor optical amplifier-based nonlinear optical loop mirror

We propose and demonstrate a scheme to implement photonic multi-shape ultra-wideband(UWB) signal generation using a semiconductor optical amplifier(SOA) based nonlinear optical loop mirror(NOLM).By employing the cross phase modulation(XPM) effect,cross gain modulation(XGM),or both,multi-shape UWB waveforms are generated including monocycle,doublet,triplet,and quadruplet pulses.Both the shapes and polarities of the generated pulses are flexible to adjust,which may be very useful in UWB pulse shape modulation and pulse polarity modulation.     

On the feasibility of full pattern-operated all-optical XOR gate with single semiconductor optical amplifier-based ultrafast nonlinear interferometer

The possibility of implementing an ultrafast all-optical XOR gate using a single semiconductor optical amplifier (SOA)-based ultrafast nonlinear interferometer (UNI) is theoretically investigated and demonstrated. For this purpose a comprehensive model that characterizes the performance of a SOA when it is successively driven by two strong pseudorandom binary sequences is applied to simulate the specific module under dual rail switching mode of operation. In this manner an extensive set of curves is obtained allowing to analyze and evaluate the impact of the input data, SOA and interferometer critical parameters on the fully loaded Q-factor. Their thorough study and interpretation reveals that the satisfaction of their requirements in order to render acceptable this metric is feasible from a technological perspective and thus if their selection is made according to the extracted guidelines then pattern-free and error-free modulo-2 arithmetic can be straightforwardly realized at 20 Gb/s. This prediction can be of practical interest in simplifying and assisting the design of more sophisticated interconnections of enhanced combinatorial and sequential functionality in which the XOR gate is the core logical unit.     

Ultrawideband monocycle generation using cross-phase modulation in a semiconductor optical amplifier

We propose a novel scheme to generate ultrawideband (UWB) monocycle pulses based on cross-phase modulation (XPM) of a semiconductor optical amplifier (SOA). The proposed system consists of a SOA and an optical bandpass filter (OBF). Due to the XPM, a continuous wave (CW) probe signal is phase modulated by another optical Gauss pulse in the SOA. The OBF will convert the phase modulation to intensity modulation. A pair of polarity-reversed UWB monocycle pulses is achieved by locating the probe carrier at the positive and negative linear slopes of the OBF. Both cases conform to the UWB definition of the Federal Communications Commission.