An effective numerical method for gain profile optimizations of multi pumped fiber Raman amplifiers

In this paper, we have solved propagation equations of multi-pump fiber Raman amplifier using Runge–Kutta (RK 4th order) numerical method and pump power evolutions along with the fiber length. They are used to calculate the net gain and gain ripple by varying the input signals powers for different fiber lengths. The pump powers are optimized by genetic algorithm and resulting net gain and gain ripple are reported graphically as well as in tabular form. The optimum minimum gain ripple is 0.26 dB for 1 mW input signal powers for 50 km fiber length. By increasing the fiber length gain ripple increases to 0.5 dB for 0.1 mW input signal power. In comparison to other methods reported in the literature, our method is simple to implement and efficient for numerical design of Raman amplification in optical communication systems.     

Gain ripple minimization in fiber Raman amplifiers based on variational method

In this paper, the variational method is employed for minimizing the gain ripple of multi-wavelength fiber Raman amplifiers. The variance of gain spectrum of the fiber Raman amplifier is regarded as the cost function, restriction on total pump power and average gain is given as the constraints of the minimization problem. It is shown that the minimization problem with any necessary constraints on the pump powers, average gain and signal to noise ratio, is reduced to a two-point boundary value problem. The method gives the entire possible local and global solutions. The method is applied to different examples of fiber Raman amplifiers with different lengths from 25 km to 100 km and different numbers of pumps from 4 to 20 to determine the pump powers and wavelengths for minimum gain ripple. It was obtained for a 100 km fiber Raman amplifier the gain ripple can be about 0.1 dB with on-off gain more than 20 dB.     

Optimization of Raman gain spectrum using Lagrangian multiplier method

Based on Lagrangian multiplier method (LMM), the gain spectrum of distributed multi-pump Raman amplifier (DMRA) is optimized. Beside of governing rate equations in Raman amplifiers, we consider two applicable constrains on cost function for optimizing the gain spectrum. Because of more accurate and controllable designing of DMRAs, a maximum power for each input pump wave is considered as well as average gain value as two constrains. LMM reduces the optimization problem with constrains to solve a two-point boundary value problem for each solution. This method is applied to minimizing gain ripple of full bandwidth (60 nm) a 100 Km Raman amplifier for different number of pumps.     

Flat gain spectrum design of Raman fiber amplifiers based on particle swarm optimization and average power analysis technique

In this work, the modified particle swarm optimization is used as an optimization tool to determine the set of wavelengths and power levels of pumps that delivers a flat gain spectrum for Raman fiber amplifiers. The average power analysis technique is used as a numerical method to solve the coupled Raman amplifier equations. By combining the modified particle swarm optimization and average power analysis technique an efficient algorithm for the design of flat-gain-spectrum broadband Raman fiber amplifiers is constructed. Application of this algorithm to the design of flat-gain-spectrum broadband Raman fiber amplifiers shows that the design efficiency of the new method is improved by 1-2 orders of magnitude compared with similar implementations previously reported in the literature. A 4-backward-pump gain-flattened Raman amplifier with bandwidth of 100-nm and maximum gain ripple of <1.0 dB is designed to demonstrate the technique.     

Characteristics of low noise hybrid fiber amplifier

The characteristics of hybrid fiber amplifier (HFA) are investigated. HFA is composed of three stages: short-length EDFA pre-stage, DCF Raman amplifier, and power boosting EDFA. HFA has low noise figure, high output power, and also wide input power dynamic range. Gain control method of HFA is presented experimentally, and the transient gain excursion is suppressed to less than 0.5 dB at 3 dB channel add-drop. HFA can be used as line amplifier in optical transmission link even combined with distributed Raman amplifier due to wide input power dynamic range. The transmission performance of HFA is better than EDFA by more than 1.0 dB of Q-factor in 720 km SMF transmission.     

Shooting algorithm and particle swarm optimization based Raman fiber amplifiers gain spectra design

An initial value determination method with a contraction factor for the counter-pumped Raman coupled equations is proposed. This method is used in conjunction with initial guess correction mechanism of Newton's method to construct a new efficient shooting algorithm for the solution of counter-pumped Raman coupled equations. The particle'swarm optimization is used to find the optimal wavelengths and powers for the pumps. By combining the new shooting algorithm and particle swarm optimization a powerful approach to the design of gain spectra for Raman fiber amplifiers is developed. Using this approach a counter-pumped broadband Raman fiber amplifier in C + L-band is designed and optimized. An average on-off gain of 9.3 dB for a bandwidth of 95 nm is obtained using only 4 pumps, with an in-band ripple level of ± 0.7 dB.     

Broad-band EDFA gain flattening by using an embedded long-period fiber grating filter

An erbium-doped fiber amplifier (EDFA) gain flattening technique using an embedded long period grating (ELPG) is proposed. By bending the ELPG, due to different coupling strengths yielded from different bending curvatures, it can be used for both the static and dynamic gain flattening despite of the different pump currents of the EDFA. The experimental results demonstrate that the flattened gain region of 34 nm can be achieved within 1 dB ripple.     

Automatic gain control in Raman amplifier with multi-wavelength pumps

Automatic gain control method in Raman amplifier with multi-wavelength pumping scheme is presented. Monitoring of several channel power and feedback pump control is used in the gain control method. The condition to minimize the gain deviation is investigated by numerical simulation. Two monitoring channels are necessary to confine gain deviation in ±0.2 dB at two pumps C-band Raman amplifier. In the experiment, gain deviation of 1.5 dB is controlled to 0.2 dB at 38/40 channels drop, and fast suppression of transient gain excursion is achieved.     

Multi parameter optimization of Raman Fiber Amplifier using genetic algorithm for S band dense wavelength division multiplexed system

A simple genetic algorithm is implemented to perform multi parameter optimization of Raman Fiber Amplifier for 100 channel S band dense wavelength division multiplexed system at 25 GHz interval. A cost effective system using single Raman pump is investigated aiming at maximum average gain. The single counter propagating pump is optimized to frequency of 211.528 THz and 652.93 mW power level with optimum Raman fiber length of 44.064 Km. There is evidence to show that the optimum solution presents a small gain variation (less than 3 dB) over an effective bandwidth covering 197–199.475 THz. The optimized configuration enabled an adequate system performance in terms of acceptable Q-factor (19.52 dB) and BER (1.46 × 10⁻²¹).     

High sensitivity of palladium on porous silicon MSM photodetector

In this work, the nanocrystalline porous silicon (PS) is prepared through the simple electrochemical etching of n-type Si (1 0 0) under the illumination of a 100 W incandescent white light. SEM, AFM, Raman and PL have been used to characterize the morphological and optical properties of the PS. SEM shows uniformed circular pores with estimated sizes, which range between 100 and 500 nm. AFM shows an increase in its surface roughness (about 6 times compared to c-Si). Raman spectra of the PS show a stronger peak with FWHM=4.3 cm⁻¹ and slight blueshift of 0.5 cm⁻¹ compared to Si. The room temperature photoluminescence (PL) peak corresponding to red emission is observed at 639.5 nm, which is due to the nano-scaled size of silicon through the quantum confinement effect. The size of the Si nanostructures is estimated to be around 7.8 nm from a quantized state effective mass theory. Thermally untreated palladium (Pd) finger contact was deposited on the PS to form MSM photodetector. Pd/PS MSM photodetector shows lower dark (two orders of magnitude) and higher photocurrent compared to a conventional Si device. Interestingly, Pd/PS MSM photodetector exhibits 158 times higher gain compared to the conventional Si device at 2.5 V.     

Performance of analog optical links employing fiber Raman amplifiers

We present the characterization of an analog optical link employing external modulation and optical carrier in the C-band using a fiber Raman amplifier (FRA). The figures of merit gain and noise figure were characterized from 1 to 9 GHz, whereas the spurious-free dynamic range was measured and compared for both setups for frequencies from 2 to 9 GHz, all with and without the Raman amplifier. Experimental results demonstrate that FRA can simultaneously increase the spurious-free dynamic range, the analog gain and decrease the analog noise figure when compared to the same configuration without optical amplifier.     

Enhanced Raman amplifier flatness with nonlinear broadening over non-standard transmission fibre

Nonlinear CW pump broadening over non-standard transmission fibre is used for the first time to achieve improved gain flatness in a single-pump broadband Raman amplifier. As an illustration of the benefits that can be obtained from this approach, a threefold increase in the bandwidth for 0.1 dB gain variation is reported when the broadened pump is used to produce 9.2 dB on-off gain over 25 km LEAF fibre.     

Optimization of pump parameters for gain flattened Raman fiber amplifiers based on artificial fish school algorithm

In this work, a novel metaheuristic named artificial fish school algorithm is introduced into the optimization of pump parameters for the design of gain flattened Raman fiber amplifiers for the first time. Artificial fish school algorithm emulates three simple social behaviors of a fish in a school, namely, preying, swarming and following, to optimize a target function. In this algorithm the pump wavelengths and power levels are mapped respectively to the state of a fish in a school, and the gain of a Raman fiber amplifier is mapped to the concentration of a food source for the fish school to search. Application of this algorithm to the design of a C-band gain flattened Raman fiber amplifier leads to an optimized amplifier that produces a flat gain spectrum with 0.63 dB in band ripple for given conditions. This result demonstrates that the artificial fish school algorithm is efficient for the optimization of pump parameters of gain flattened Raman fiber amplifiers.     

分布式拉曼光纤放大器动态增益控制研究

介绍了动态增益控制的必要性，对动态拉曼传输方程进行了简化，并将其以矩阵形式表示，从而减少了方程的数量，提高了计算速度。由拉曼传输耦合方程推出一种适用于分布式拉曼放大器实时控制的自动控制算法。考虑工程需要，该算法忽略了噪声功率、泵浦间的受激拉曼效应，以及信号和泵浦间受激拉曼效应对泵浦功率的损耗。结果表明该算法能够达到快速抑制输入信号功率突变引起的输出功率／增益波动的目的。     

Effect of sinusoidal ripples in refractive index profile distribution on the performance characteristics of dual concentric core step index fiber Raman amplifier

Phased matched wavelength, effective area, effective Raman gain, and wave guide dispersion are computed from exact numerical solution assuming scalar wave equation in the presence and absence of ripples as imperfections in the refractive index profile in dual cores of single mode fiber Raman amplifier for the first time. It is observed that for larger values of amplitude and lower frequencies, the effective Raman gain increases w.r.t. that calculated with no ripples. However, we assume ripple amplitude up to 5% of core cladding refractive index difference w.r.t. the available data, corresponding to the three ranges of relative ripple amplitudes of 1%, 2%, 3% and ripple frequencies of 1, 2, 3 μm⁻¹. Based on these data, we analyse performance of FRA over frequency shift band of 20–700 cm⁻¹. Uniformity of gain is interestingly seen to be maintained for higher ripple frequency and lower amplitude. However, no prominent effect in coefficient of dispersion and phase matched wavelength is observed within operating range of wavelength. Also, based on available structural parameter, the investigation should find use as a guide to system users to know the limit and promise of existence of ripples.     

L-band all-optical gain-clamped EDFA by utilizing C-band backward ASE

By using an optical circulator and C/L-band wavelength division multiplexer to recycle the C-band backward ASE, an L-band gain-clamped erbium-doped fiber amplifier is presented. We have experimentally studied the static gain clamping property of this amplifier. As the ASE feedback attenuation is set to 0, the gain at 1585 nm can be clamped at 18.84 ± 0.26 dB within dynamic range of 25 dB and the critical power reaches about −15.09 dBm. The gain variation and saturated output power at 1585 nm for 0 dB attenuation are 1 dB lower and 2.17 dB higher than those for 30 dB attenuation, which indicates that the L-band EDFA gain can be effectively clamped via the ASE injection technique.     

Two-wavelength lidar inversion algorithm for determination of aerosol extinction-to-backscatter ratio and its application to CALIPSO lidar measurements

A modified two-wavelength lidar inversion algorithm is proposed to aid in the retrieval of aerosol extinction-to-backscatter ratios (lidar ratio) as well as backscatter coefficients and extinction color ratios from simultaneous two-wavelength elastic backscatter lidar measurements. To demonstrate the feasibility of the algorithm, both the Raman method and the two-wavelength method have been applied to the ground-based measurements at 355 and 532 nm; moreover, it has been applied to the data acquired by the Cloud Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) lidar, and to the simultaneous ground-based lidar measurements carried out at Napoli (southern Italy, 40.838 °N, 14.183 °E, 118 m above sea level). Three cases of Saharan dust transport towards Europe have been considered. From the comparison, it can be found that the values of lidar ratio and backscatter coefficient retrieved by the modified two-wavelength algorithm are in good agreement with those obtained by the Raman method. Moreover the retrieved mean values of the lidar ratios and color ratios are in agreement with those reported by other authors.     

Surface plasmon resonance in near-field coupled gold cylinder arrays fabricated by EUV-interference lithography and hot embossing

Arrays of metal nanoparticles with nanometer-scale gaps between the particles is highly interesting for plasmonic field enhancement applications. We report a simple method to fabricate arrays of closely spaced Au particles with inter-particle separation down to 20 nm. We used extreme ultraviolet interference lithography (EUV-IL) and a mechanical press to fabricate two-dimensional arrays of Au nanoparticles. Lithographically produced particle arrays were modified by hot pressing in a nanoimprint machine and the gap was varied in a range from 50 nm to below 20 nm. Optical measurement shows two resonances at 520 nm and 620 nm, with the latter gaining strength as the gap is reduced. The experimental and theoretical investigations using a FDTD algorithm demonstrate that the low-energy resonance can be assigned to a collective surface plasmon resonance arising from the strong near-field coupling between the nanoparticles. Surface Enhanced Raman Spectroscopy (SERS) experiments performed on a model molecule (BPE) show a large gain in signal intensity as a result of the reduced gaps between the particles.     

Gain and gain-flatness improved photonic crystal fiber Raman amplifier based on a dual-pass amplification configuration with single pump laser

A gain and gain-flatness improved L-band dual-pass Raman fiber amplifier (RFA) utilizing a photonic crystal fiber (PCF) as gain medium is demonstrated. By introducing complementary gain spectra of typical forward and backward pumping single-pass RFA using the same PCF, we finally achieve average net gain level of 22.5 dB with a ±0.8 dB flattening gain in 20-nm bandwidth from 1595 nm to 1615 nm, which is rare in RFAs with only one single pump and no flattening filter. Compared with the single-pass pump configurations, gain level, flatness and bandwidth are greatly improved by using the dual-pass amplification configuration. The limitation of this configuration caused by multi-path interference (MPI) noise and stimulated Brillouin scattering (SBS) is also discussed.     

Thermally tunable EDFA gain equalizer using point symmetric cascaded Mach-Zehnder coupler

In this paper, a thermally tunable EDFA gain equalizer filter based on point symmetric cascaded Mach-Zehnder (CMZ) filter based two mode interference (TMI) coupler is presented with its mathematical model. Transmission characteristics of these CMZ couplers are analyzed and compared with Y symmetric CMZ couplers by using this model. For EDFA gain equalizer, point symmetric CMZ circuit is chosen due to its higher wavelength flattening width than Y symmetric CMZ circuit. The ripples of equalized EDFA gain spectrum are formulated and estimated from the equalized gain spectrum of point symmetric CMZ filters. It is found that 2 stage point symmetric CMZ coupler with binomial coupler distribution (2PB CMZ) using Δn = 5% provides gain equalized width of 35 nm with ripple of 0.4-0.6 dB and bending loss of 0.24 dB and device length is ∼15 times lower than that of the existing EDFA gain equalizer based CMZ filter. It is also seen that if during the fabrication process, waveguide core width w is increased or decreased by 0.1 μm (in percentage ∼±6.6%), the power imbalance of TMI based 2PB CMZ filter is slightly increased by ∼8% in comparison to that based on directional coupler (DC) by 40%. Low power thermooptic structure of varying gap between two waveguide cores with silicon trench just below the heater is used and it requires ∼1.5 times less heating power than the conventional structure for thermal tuning of EDFA gain equalization.