Dual-ring 16-Star QAM direct and coherent detection in 100 Gb/s optically amplified fiber transmission: simulation

This paper presents the transmission performances, by simulation, of optical communication systems over 10 spans of dispersion compensated and optically amplified fiber distance with a bit rate of 100 Gb/s, by employing modulation formats of two amplitude levels and 8 phase states per amplitude level, the 2R-16-Star QAM constellation under direct and coherent detection with and without phase estimation. Different ring ratios of the amplitude levels are examined and associate transmission performances are reported. Optical signal to noise ratio is achieved with 18 dB and 23 dB for back to back and long haul transmission cases with a dispersion tolerance of  ±67 ps/nm at 2 dB power penalty of the eye opening at 100 Gb/s. Monte–Carlo simulation is also performed and a receiver sensitivity of  − 15 dBm is achieved for a BER of 10⁻⁵under direct detection after 1100 km of dispersion-compensated and optically amplified transmission. Transmission performances, bit error rate versus receiver sensitivity, are also confirmed with the use of the eye diagram and associate multiple-peaks statistical spectral density distribution. For 100 Gb/s 2R-16-Star QAM coherent transmission, an improvement of the receiver sensitivity of 2.5 dB and 3.5 dB is obtained for coherent detection without phase estimation and respectively. Under coherent detection with phase estimation, the chromatic dispersion tolerance reaches  ±100 ps/nm for a 2 dB eye opening penalty at 100 Gb/s bit rate. Comparative studies of the transmission performances of the Star and Square QAM modulation formats are also conducted under the fiber linear and nonlinear effects and detection with and without phase estimation.     

The effects of external optical feedback on the power penalty of DFB-LD modules for 2.5Gbs−1 optical transmission systems

We report the effects of external optical feedback on the power penalty of commercial distributed feedback laser diode (DFB-LD) modules for 2.5Gbs–1 optical transmission systems. External optical feedback presented to the DFB-LD modules causes the excitation of external cavity modes, resulting in increased relative intensity noise (RIN) and intensity noise ripples at low frequency region below 500MHz. For a 10–10 bit error rate (BER), the minimum power penalty is as much as 1.25dB for a feedback ratio of –8.8dB. An excess power penalty of 0.5dB per 3dB increase in the feedback ratio was also empirically obtained. We suggest that optical isolators in 2.5Gbs–1 DFB-LD modules used in conventional optical transmission systems or WDM systems must have a peak isolation ratio of better than 54.5dB, instead of the previously recommended 30dB, for negligible power penalty induced by external optical feedback.     

Simulation of DWDM signals using optimum span scheme with cascaded optimized semiconductor optical amplifiers

We numerically simulated the ten channels at 10 Gb/s dense wavelength division multiplexing (DWDM) transmission faithfully over 17,227 km using 70 km span of single mode fiber (SMF) and dispersion compensating fiber (DCF) using optimum span scheme at channel spacing 20 GHz. For this purpose, inline optimized semiconductor optical amplifiers (SOAs) and DPSK format are used. We optimized the SOA parameters for inline amplifier with minimum crosstalk and amplified spontaneous emission noise with sufficient gain at bias current 400 mA. For this bias current, constant gain 36.5 dB is obtained up to saturation power 21.35 mW. We have also optimized the optical phase modulator bandwidth for 400 mA current which is around 5.5 GHz with crosstalk −14.2 dB between two channels at spacing 20 GHz.We show the 10×10 Gb/s transmission over 70 km distance with inline amplifier has good signal power received as compared to without amplifier, even at equal quality factor. We further investigated the optimum span scheme for 5670 km transmission distance for 10×10 Gb/s with channel spacing 20 at 5.5 GHz optical phase modulator bandwidth. As we increase the transmission distance up to 17,227 km, there is increase in power penalty with reasonable quality.The impact of optical power received and Q factor at 5670 and 17,227 km transmission distance for different span schemes for all channels has been illustrated. For launched optical power less than saturation, all channels are obtained at bit error rate floor of 10⁻¹⁰.     

Minimization of Cross-Gain Saturation in Wavelength Division Multiplexing by Optimizing Differential Gain in Semiconductor Optical Amplifiers

We successfully simulated the 10 × 40 Gbit/s soliton RZ-DPSK WDM signals over 1050 km with spectral efficiency approaching 0.4 bit/s/Hz using semiconductor optical amplifiers (SOAs) as in-line amplifier. The cross-gain saturation of SOA can be minimized by settling crosstalk at a lower level by decreasing the differential gain. This decrease in differential gain is in such a way that we get nil power penalty. The maximum transmission distance of 1050 km is possible with differential gain 210 atto cm² of SOA.

The impact of amplification factor, ASE noise power, crosstalk, quality factor and bit error rate for different differential gain has been investigated. It has been shown that with the increase in differential gain of SOA, the transmission distance goes on decreasing. At high value of differential gain 2.5 × 10^-16 cm² for the transmission distance 1050 km, all channels produce inter channel crosstalk with bit error rate greater than 10^-6. But for lower differential gain 190 atto cm², the quality of all channel increases at the cost of large power penalty.

With slight increase in differential gain 200 atto cm², the maximum transmission distance observed is 4550 km with quality of received signal more than 15 dB and having nil power penalty. We observed clear eye diagrams and optical power spectra for received signal with transmission distance 1050 km and 4550 km using soliton RZ-DPSK system. The bit error rate for all channels increase more than 10^-10 with the increase in launched input power that is due to power saturation.     

A terabit capacity passive polymer optical backplane based on a novel meshed waveguide architecture

An optical backplane based on a meshed polymer waveguide architecture enabling high-speed board-to-board optical interconnection is presented. This planar array of multimode polymer waveguides can provide passive strictly non-blocking links between server line cards fitted with optical transmitter and receiver arrays. This architecture offers a scalable and low-cost solution to the bandwidth limitations faced by electrical backplanes and is suitable for PCB integration. The reported backplane demonstrator uses a matrix of 100 waveguides each capable of 10 Gb/s operation to interconnect 10 cards for a total capacity of a terabit per second aggregate data rate in multicast mode. Characterisation of the backplane demonstrator reveals low link losses of 2 to 8 dB for a multimode fibre input and crosstalk values below −35 dB. Error free data transmission at 10 Gb/s is achieved with a power penalty of only 0.2 dB at a bit-error-rate of 10⁻⁹. Additionally, lossless operation of a Gigabit Ethernet link over the backplane is achieved even when using the worst-case highest loss links.     

Add/drop applications in fiber ring networks based on a reconfigurable optical add/drop multiplexer in a re-circulating loop

A periodic add/drop system in a fiber ring network was investigated using a reconfigurable optical add/drop multiplexer (ROADM) in a re-circulating fiber loop. After seven cascaded add/drop nodes at every 150 km along the transmission, at bit error ratio (BER) equals to 10⁻⁹ and data rate of 10 Gbps, we observed a 2.5 dB power penalty for the passing through channels with 1050 km transmission distance, and 0.3 dB sensitivity penalty variation for the periodic add/drop channels at every 150 km, respectively.     

Y-branch - the key component for crosstalk reduced optical switches: modelling analysis

A novel design method for switching devices is demonstrated based on waveguide structures with Y-branches and abrupt waveguide changes as key elements. A combination of BPM calculation and eigenmode analysis is shown to be a promising tool to obtain deep insight into the physical mode behaviour. Based on the analysis, a mode converter to match the eigenmodes in amplitudes and phase can be developed and introduced as a correcting device. This method is demonstrated for the crosstalk compensation of a digital switch (DOS) and a Δβ-reversal switch. A crosstalk of about −20 dB can be reduced to −40 dB by this method. This revised version was published online in November 2006 with corrections to the Cover Date.     

Study of gain flatness for multi-channel amplification in single stage EDFA for WDM applications

In this paper, gain flatness is studied for simultaneous 16-ITU-T channel amplifications at C-band (1,532–1,558 nm) in a single stage EDFA for WDM application at different average inversion levels. The inversion levels are varied due to the change of the input signal levels from the targeted operating point and also for dropping few numbers of channels. Specially designed gain flattening filter (GFF) is used in order to get the flat gain with gain variation ±0.5 dB for −20 dBm/ch input signal power (total input signal power is −8.0 dBm) at a fixed average inversion level which is maintained by proper selection of optimum fibre length and pump power. A specific loss spectrum of GFF is obtained by writing a chirped fibre Bragg grating of length 20 mm. Gain variations are studied by changing the total input signal levels from −8.0 dBm to −20.0 dBm and maintained within 20.0 ± 0.5 dB by using automatic gain control (AGC) circuit. About 15 out of 16 channels are dropped and observed ± 0.5 dB gain-variation which is an important parameter in optical network system.     

Performance analysis of synchronous coherent optical systems in the presence of multiple in-line amplifiers

The performance of synchronous binary and M-ary coherent optical systems with cascaded optical amplifiers are analysed. From the relative performance of the various demodulation techniques that are compared in terms of power penalties due to amplifier input power, noise accumulation and bit rate, it is predicted that binary homodyne PSK and M-ary FSK schemes not only accommodate more noise accumulation due to multiple amplifiers but also facilitate improvement in dynamic range. The results on the performance of systems in the 1–10Gb/s range show that total system lengths of the order of 5000km can be achieved at 10Gb/s for 16-FSK and homodyne PSK for a bit error rate of 10−9 and 1 dB power penalty, whereas it is reduced by a factor of about two for DPSK which was the best of the direct detection and asynchronous coherent schemes that have been previously analysed. This revised version was published online in November 2006 with corrections to the Cover Date.     

An integrated reconfigurable optical add-drop multiplexer based on using an opto-VLSI processor and a 4-f imaging system

A novel integrated reconfigurable optical add-drop multiplexer (ROADM) structure is proposed and demonstrated experimentally. The ROADM employs an interface substrate that enables the integration of a fiber-lens collimator array and an Opto-VLSI processor to realise a 4-f imaging system through optical beam steering. Phase hologram optimization of the Opto-VLSI processor is investigated, for maximizing the number of wavelength channels while keeping adequate insertion loss and crosstalk levels. Experimental results are shown, which demonstrate a broadband operation and the principle of the ROADM with crosstalk of less than −30.5 dB and insertion loss of less than 6 dB for the drop and through operation modes.     

A detailed analysis of cross-phase modulation effects on OOK and dpsk optical WDM transmission systems in the presence of GVD,SPM, and ASE noise

A performance analysis is carried out to evaluate the effect of cross-phase modulation (XPM) on a dispersion-managed 20 Gb/s optical wavelength-division multiplexing (WDM) transmission system using either the on-off keying (OOK) or the different-phase-shifting keying (DPSK) modulation, in the presence of the group-velocity dispersion (GVD), self-phase modulation (SPM), and amplified spontaneous emission (ASE). It is found that to achieve a bit error rate (BER) of 10⁻⁹ at a distance of 160 km, a 1.0 dB XPM power penalty is incurred for input channel power of 3 dBm in the OOK transmission and 7 dBm in the DPSK transmission. The power penalty increases with input channel powers and is inversely proportional and exhibits oscillations with respect to the channel separation. The oscillation is evenly spaced for the DPSK but not for the OOK and suggests the presence of optimum separation values. The XPM penalty decreases when a high dispersion fiber is used and increases linearly with increasing dispersion slope. Small residual dispersion can reduce the penalty of nonlinear effects.     

SOA nonlinearities in 4 × 25 Gb/s WDM pre-amplified system for 100-Gb/s Ethernet

A simulation analysis of the impact that the nonlinear response of a semiconductor optical amplifier (SOA) has in a four-channel WDM pre-amplified transmission system is presented in the framework of the recently proposed extended-reach (40-km, 4 × 25 Gb/s) 100 Gb Ethernet link. Channel spacing values ranging from 200 to 800 GHz, and fiber losses between 0 and −20 dB are considered. A maximum power penalty of 4.5 dB is predicted for short fiber lengths and for the tightest channel plan. For short fiber lengths, the penalty drops by about 0.8 dB when moving from 400 to 800 GHz; whereas for long fiber lengths, the penalty increases by 0.2 dB, provided that an average dispersive fiber is utilized. The widely spaced channel plan then represents the best choice in terms of the analyzed physical effects to implement the next-generation 100 GbE link. Further, our numerical investigation includes a discrimination analysis that confirms cross-gain modulation as the main overall SOA nonlinear impairment in the analyzed architecture, and establishes ultra-fast carrier heating-induced FWM as responsible for the system performance difference observed as a function of channel spacing. The difference practically vanishes for fiber lengths above 30 km. Finally, the proposal of an equation that fits the simulated power spectrum density of the first-order four-wave mixing-generated product as a function of channel spacing is presented as an aid to validate our numerical results.     

Multichannel dispersion compensation using a silicon waveguide-based optical phase conjugator

We experimentally demonstrate dispersion compensation using a silicon-based optical phase conjugator. We achieve simultaneous transmission of four dense wavelength division multiplexing (DWDM) channels spaced at 100 GHz and operating at 10 Gbits/s over 320 km of standard fiber. The measured power penalty at bit error rate of 10(-9) is less than 0.3 dB.     

Systematic investigation of silicon digital 1×2 electro-optic switch based on a microdisk resonator through carrier injection

We perform a systematic investigation on the silicon digital 1×2 electro-optic switch based on a microdisk resonator with the refractive index tuned by carrier injection and extraction. Analytical expressions are proposed to quantify the crosstalk, insertion loss and switching time, and then they are used to analyze the effect of the coupling-in/out coefficient, internal loss, and quality factor. Both electron beam lithography and photolithography were employed to construct the asymmetrically coupled microdisk resonator integrated with a p-i-n diode. A high-performance switch has been demonstrated with a crosstalk of less than −20 dB, a switching time of ∼2 ns, as well as a power consumption of 0.46 mW. Moreover, the switching of optical signals with high data rates has been characterized and analyzed under different working conditions.     

密集波分复用条件下的光轨网络串扰分析与仿真

光轨网络是一种能够利用成熟的光学器件实现带宽灵活分配和信息交换的新型网络.串扰是限制光轨网络的物理层性能及其扩展性的重要因素.本文讨论了典型的光轨网络节点中异频串扰和同频串扰的产生原因,理论分析了两者对光轨网络的物理层传输性能的影响.给出了3种串扰性能的评价方法.以密集波分复用技术为应用背景,分别搭建了器件隔离度为20dB和30dB的、具有3个节点5个波长且单波长速率为2.5Gbps的光轨网络,仿真了串扰在光轨网络中的传播过程,并计算了光轨网络的误码率、功率代价和相对串扰.理论分析和仿真结果表明:光滤波器、解复用器和复用器是光轨网络中串扰产生的关键器件,且提高器件的隔离度等性能对于提高光轨网络的传输性能会有较显著的效果;在密集波分复用条件下,串扰对单波长速率为2.5Gbps的光轨网络的误码率和功率代价具有显著的影响,从而限制了光轨网络实际可用的节点数目.     

Gain clamping in double-pass L-band EDFA using a broadband FBG

A highly efficient gain-clamped L-band EDFA with improved noise figure characteristic is demonstrated by simply adding a broadband C-band FBG in double-pass system. The combination of the FBG and optical circulator has created laser in the cavity for gain clamping. By adjusting the power combination of pumps 1 and 2, the clamped gain level can be controlled. The amplifier gain is clamped at 28.1 dB from −40 to −25 dBm with a gain variation of less than 0.5 dB by setting the pumps 1 and 2 at 59.5 and 50.6 mW, respectively. The gain is also flat from 1574 nm to 1604 nm with a gain variation of less than 3 dB. The corresponding noise figure varies from 5.6 to 7.6 dB, which is 0.8 to 2.6 dB less than those of unclamped amplifier.     

Gain-flattened erbium-doped fiber amplifier with flexible selective band for optical networks

Erbium-doped fiber amplifier with flat gain over 30 nm bandwidth is demonstrated using flexible selective band methods. The band optical amplifier was designed to cater 44 wavelength division multiplexing channels which were separated into bands of 4 nm. Without using any gain flattening filter, the gain of optical amplifier was maintained at 19 dB with a maximum gain variation of less than 1.6 dB even though the input signal power was varied from −19 to −6 dBm. The amplifier was able to maintain 1 dB gain flatness with 83% chance for any selective bands of 4 nm within the wavelength range from 1530 to 1565 nm. This feature is very attractive to support band optical networks.     

WDM光网络串扰特性分析与实验研究

提出了WDM光网络系统串扰分析统一的数学模型,分析比较了WDM光网络系统各种串扰的起源与特征,实验研究了不同的非相干串扰代价下,WDM光网络系统的功率代价.对应1dB功率代价下,得出WDM光网络系统对应非相干串扰代价容限小于-20dB的结果.     

基于空间光调制器的MIMO大气传输特性研究

大气湍流对无线光通信系统的影响不可忽视,为了更准确地反映实验室模拟多输入、多输出(MIMO)大气湍流信道的实际特征,提出了一种利用相位屏来模拟MIMO大气湍流信道的方法,并针对基于液晶空间光调制器（LC-SLM）的液晶调制法展开研究,通过实验验证该方法的可行性。实验结果表明：通过相位屏模拟MIMO大气湍流信道的激光光斑发生不同程度的畸变,湍流环境下两路激光发射系统比单路发射激光系统功率稳定性好,在前向纠错误差极限（3.8×10-3）下,单个发射单个接收系统的链路代价为10.5 dB,2个发射2个接收的MIMO系统的链路代价为9.3 dB。该项研究对于实验室模拟MIMO大气湍流信道实验方法提供一种新思路。     

密集波分复用条件下的光轨网络串扰分析与仿真

光轨网络是一种能够利用成熟的光学器件实现带宽灵活分配和信息交换的新型网络.串扰是限制光轨网络的物理层性能及其扩展性的重要因素.本文讨论了典型的光轨网络节点中异频串扰和同频串扰的产生原因,理论分析了两者对光轨网络的物理层传输性能的影响.给出了3种串扰性能的评价方法.以密集波分复用技术为应用背景,分别搭建了器件隔离度为20 dB和30 dB的、具有3个节点5个波长且单波长速率为2.5 Gbps的光轨网络,仿真了串扰在光轨网络中的传播过程,并计算了光轨网络的误码率、功率代价和相对串扰.理论分析和仿真结果表明:光滤波器、解复用器和复用器是光轨网络中串扰产生的关键器件,且提高器件的隔离度等性能对于提高光轨网络的传输性能会有较显著的效果;在密集波分复用条件下,串扰对单波长速率为2.5 Gbps的光轨网络的误码率和功率代价具有显著的影响,从而限制了光轨网络实际可用的节点数目.