Simulation and optimization of a polymer directional coupler electro-optic switch with push-pull electrodes

Structural model and design technique are proposed for a polymer directional coupler electro-optic switch with rib waveguides and push-pull electrodes, of which the electric field distribution is analyzed by the conformal transforming method and image method. In order to get the minimum mode loss and the minimum switching voltage, the parameters of the waveguide and electrode are optimized, such as the core with, core thickness, buffer layer between the core and the electrode, coupling gap between the waveguides, electrode thickness, electrode width and electrode gap. Switching Characteristics are analyzed, which include the output power, insertion loss, and crosstalk. To realize normal switching function, the fabrication error, spectrum shift, and coupling loss between a single mode fiber (SMF) and the waveguide are discussed. Simulation results show that the coupling length is 3082 μm, push-pull switching voltage is 2.14 V, insertion loss is less than 1.17 dB, and crosstalk is less than −30 dB for the designed device.     

EAM-SOA millimeter-wave frequency up-converter for radio-over-fiber applications

We present an optical scheme for photonic frequency up-conversion at the millimeter-wave bands based on Semiconductor Optical Amplifier. The proposed scheme modulates the bias current with the Intermediate Frequency in order to achieve frequency mixing of an incoming optical signal modulated with the Local Oscillator. Theory indicates that the proposed scheme supports data bandwidths in the tens of GHz for LO values above 10 GHz. This scheme allows for photonic integration of the considered optical devices. A laboratory demonstration of the scheme for up-conversion to the 40 GHz band, using narrow-band IF signals, showed relatively low thresholds for the optical input power and bias current level to achieve error free operation: − 14.5 dBm 100 mA for a 64-QAM signal. Spurious-Free Dynamic Range showed an acceptable performance, with a linearity about 52.5 dB·Hz^2/3 for an optical input power of − 6 dBm.     

Filter-free ultrawideband generation based on semiconductor optical amplifier nonlinearities

A filter-free scheme for ultrawideband (UWB) generation with single semiconductor optical amplifier (SOA) is proposed and demonstrated. A pair of polarity-reversed optical pulses is generated due to cross gain modulation (XGM) in the pump-probe scheme, whereas the amplified pump pulse becomes sharp at its leading edge such that the power peak is shifted forward. Hence, the combination of the pump and probe signals at the SOA output is quasi-monocycle shape. Our scheme is an improved scheme to avoid exploiting optical filters and time-delay devices. The generated UWB radio frequency spectrum shows good stability when the input probe power varies from −10 dBm to 6 dBm, the probe wavelength varies in the whole C-band, and the bias current varies from 100 mA to 240 mA. Although the generated quasi-monocycle deviates from a standard Gauss monocycle to some extent, the frequency spectrum conforms to the UWB regulation. Two SOAs with different XGM dynamics are compared in generating UWB signals.     

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.     

An optical power-equalized amplifier with a wide dynamic range

An optical power equalization amplifier with a wide dynamic range is proposed and demonstrated with no electronic control. It shows constant and equalized outputs when a power difference between input channels and a total input power are changed. It has more than a 15 dB dynamic range for input signals between −30 dBm and −5 dBm. The structure of this amplifier can be more promising when it is applied to a planar waveguide device.     

Highly birefringent photonic crystal fiber polarization splitter made of soft glass

A soft glass dual core polarization splitter based on highly birefringent photonic crystal fiber (PCF) is proposed and the full vector finite element method (FEM) is employed to analyze the impacts of structural parameters on birefringence and the coupling length, and simulation results show that high birefringence on the order of 10⁻² can be obtained at 1.55 μm, moreover, hole size, hole pitch and elliptic ratio all affect birefringence and the coupling length. Based on these results, the PCF's structure is optimized to realize a polarization splitter of 282 μm whose largest extinction ratio is around −45.42 dB at 1.55 μm. Meanwhile, the bandwidth at the extinction ratio of −10 dB is about 90 nm, and around 32 nm at −20 dB.     

Gain-clamped dual-stage L-band EDFA by using backward C-band ASE

A dual-stage L-band gain-clamped erbium-doped fiber amplifier (GC-EDFA) by using backward C-band amplified spontaneous emission (ASE) is proposed. Compared with other similar GC-EDFAs, the proposed structure has higher and flatter clamped gain in L-band because of its optimal pump power and EDF length. The flatness from 1570 nm to 1600 nm arrives 0.77 dB, the bandwidth of 3 dB is more than 35 nm and the maximal input signal power arrives −15 dBm.     

Design of a wavelength demultiplexer based on a parabolically tapered multimode interference coupler

A compact wavelength demultiplexer is designed for the operation at 1.30 and 1.55 μm wavelengths using the three-dimensional semi-vectorial beam propagation method. The parabolically tapered multimode interference (MMI) coupler based on the deep-etched SiO₂/SiON rib waveguide is introduced into the present demultiplexer for reducing the length. The numerical results show that a MMI section of 330.0 μm in length, which is only 76% length of a straight MMI coupler, is achieved with the contrasts of 42.3 and 39.2 dB in quasi-TE mode, and 38.4 and 37.8 dB in quasi-TM mode at wavelengths 1.30 and 1.55 μm, respectively and the insertion losses below 0.2 dB. The modified finite difference scheme is applied to approximate the resulting equations along the transverse directions, in which the discontinuities of the derivatives of magnetic field components H_y and H_x along the vertical and horizontal interfaces, respectively, are involved.     

Gain clamped L-band EDFA with forward–backward pumping scheme using fiber Bragg grating

The paper proposes a novel two stage L-band erbium doped fiber amplifier with forward–backward pumping scheme for transmission of 32 wavelength division multiplexed (WDM) channels. It is gain clamped with an in-line fiber Bragg grating (FBG) to provide flat gain over 45 nm by restricting and reutilizing amplified spontaneous emission (ASE). We demonstrate that it provides an efficient small signal gain with minimum noise figure of over 20 dB and 5.5 dB, respectively, in the L-band region (1565–1610 nm) by comparing with its forward and backward pumped counterparts with fixed Er³⁺ fiber length of 20 m for −30 dBm/channel input power. We also obtain the gain and noise figure dependence as a function of each of the Er³⁺ fiber lengths, pump power (both 1480 and 980 nm), and temperature. Hence a 10 nm region (1580–1590 nm) has been acknowledged where temperature variations become constricted for 30 °C variations (15–45 °C).     

Experimental and theoretical studies on a double-pass C-band bismuth-based erbium-doped fiber amplifier

Performance of a Bismuth-based Erbium-doped fiber amplifier is experimentally and theoretically investigated using 1480 nm pumping with double-pass scheme. In the theoretical analysis, the rate and power propagation equations are solved to examine the optimum length for the C-band operation as well as the gain and noise figure characteristics. The calculated small signal gain is 38 dB with gain variation of less than 3 dB. The measured gain is 4 dB lower due to spurious reflections which were ignored in the theoretical analysis. At input signal power of 0 dBm, a gain of 14.5 dB is obtained experimentally with gain variation of less than 1 dB within the wavelength region from 1530 to 1565 nm. The noise figure is less than 12 dB within this region.     

The L-band EDFA of high clamped gain and low noise figure implemented using fiber Bragg grating and double-pass method

The L-band erbium-doped fiber amplifier (EDFA) of low noise figure and high clamped-gain using gain-clamped and double-pass configuration is presented in this paper. A total of five different configurations of EDFAs by reflection scheme with single forward pumping schemes are examined and compared here. Among these configurations, we first find the configuration of 1480-nm pumped L-band EDFA with optimum gain and noise figure value. To further minimize the gain variation, a fiber Bragg grating (FBG) with 1615-nm center wavelength and 1-nm bandwidth is determined and added in double-pass L-band EDFA. The gain variation and maximum noise figure of EDFA while channel dropping is investigated. As the number of channel dropping from 32 to 4, the L-band type-A EDFA keep the variation of gain within 2.9 dB and the maximum noise figure below 5 dB with each channel’s input power of −23 dBm.     

Short-length and broadband mismatched optical couplers with tapered Hamming function for C- and L-band

Mismatched optical couplers with variable widths of waveguide tapered by Hamming function are numerically investigated in the demand of short-length, broadband, and low crosstalk. We used global search algorithm and beam propagation method to seek optimal structure parameters of coupling waveguide. The coupler length is 3.6 mm within the C+L-band (1.53-1.61 μm) for variable widths of waveguide at crosstalk level of −35 dB. Comparison with constant width of waveguide, the constant width of waveguide has a coupler length of 4.4 mm and can only achieve −20 dB of crosstalk within the C-band (1.53-1.565 μm). Obviously, the waveguide with variable widths has the advantage over constant width for the demand of short-length, bandwidth, and low-crosstalk.     

On duty cycle selection of RZ optical pulse to optimize the performance of dispersion compensated 10 Gbps single channel optical communication system using dispersion compensating fibers

We present results for duty cycle selection of optical RZ pulse to optimize the performance in 10 Gbps single channel dispersion compensated optical communication system. The system has link length of 240 km with two spans. Each of the spans consists of 120 km standard single mode fiber (SSMF) of 16 ps/nm/km, whose chromatic dispersion is compensated using pre-, post- and symmetrical-dispersion compensation schemes by 24 km dispersion compensating fiber (DCF) of −80 ps/nm/km. The performance of the three compensation schemes is compared by taking 8, 10, 12 and 14 dBm Er-doped fiber amplifier (EDFA) power levels in the link with a duty cycle range (0.1-0.9) of RZ optical pulse. The graphical results obtained show a relationship among the duty cycle, EDFA power and dispersion compensation scheme which predicts the best performing duty cycle case. To optimize performance of the system, we recommend in general, duty cycle less than 0.3 and EDFA power below 8 dB irrespective of compensation scheme. However, with post compensation duty cycle less than 0.7 and EDFA power below 12 dBm give optimum performance. The results conclude that for the high value of duty cycle, post dispersion compensation scheme should be used.     

Long range surface plasmon polariton waveguides at 1.31 and 1.55 μm wavelengths

We investigate characteristics of gold metal strip waveguides based on long range surface plasmon polaritons (LRSPPs) along thin metal strips embedded in a polymer for practical applications at the telecommunication wavelengths of 1.31 and 1.55 μm. Guiding properties of the gold strip waveguides are theoretically and experimentally evaluated with the limited thickness and width up to ∼20 nm and ∼10 μm, respectively. The lowest propagation loss of ∼1.4 dB/cm is obtained with a 14.5-nm-thick and 2-μm-wide gold strip at 1.55 μm. With a single-mode fiber, the lowest coupling loss of ∼0.4 dB/facet is achieved with a 14.5-nm-thick and 5-μm-wide gold strip at 1.55 μm. The lowest insertion losses are obtained 8-9 dB with 1.5 cm-long gold strips of a limited thickness and width at both the wavelengths. We demonstrate a 10 Gbps optical signal transmission via the LRSPP waveguide with a 14 nm-thick, 2.5 μm-wide, and 4 cm-long gold strip. These LRSPP waveguides have potential applications for optical interconnects and communications.     

Double-pass erbium-doped zirconia fiber amplifier for wide-band and flat-gain operations

The double-pass erbium-doped zirconia fiber amplifier (EDZFA) is proposed and demonstrated to provide a wide-band amplification as well as flat-gain operation in both the C- and L-band regions using only a single-gain medium. The proposed amplifier utilizes an erbium-doped zirconia fiber (EDZF) with erbium ion concentration of 2800 ppm as a gain medium. The medium is fabricated in a ternary glass host, zirconia-yttria-aluminum codoped silica fiber through solution doping technique along with modified chemical vapor deposition (MCVD). Compared to a single-pass operation, the double-pass EDZFA shows a better gain performance. At input signal power of 0 dBm and the optimum EDZF length of 2 m, a flat gain of around 16 dB is achieved by the proposed double-pass amplifier with gain variation of approximately 2.5 dB throughout the wavelength range from 1530 to 1590 nm. However, the noise figure of the double-pass amplifier is slightly higher than that of the single-pass due to inefficient population inversion at the input part of the amplifier.     

Experimental study of SOA-based NRZ-to-PRZ conversion and distortion elimination of amplified NRZ signal using spectral filtering

We experimentally study both reshaping of nonreturn-to-zero (NRZ) signal and NRZ to pseudoreturn-to-zero (PRZ) format conversion based on self-phase modulation of a semiconductor optical amplifier (SOA) and detuning an optical bandpass filter (OBF). When an OBF with 1 nm bandwidth is blue shifted by 0.8 nm, the distortion of the amplified NRZ signal at 10 Gbit/s is shown to be eliminated completely. When an OBF with 0.32 nm bandwidth is red shifted by 0.42 nm from the carrier frequency, NRZ-to-PRZ conversion at 10 Gbit/s is obtained. A holding beam is used to suppress the SOA noise and improve the output extinction ratio (ER). The output ER of both the reshaped NRZ and the converted PRZ is larger than 10 dB when the signal wavelength is longer than 1540 nm, and an input power dynamic range from −7 dBm to 2 dBm is obtained at a signal wavelength of 1563.6 nm. The average power of the reshaped NRZ signal is about 3 dBm at an input power dynamic range of 13 dB. The amplitude fluctuation of the converted PRZ signal is around 1.6 dB.     

Hybrid photonic crystal 1.31/1.55 μm wavelength division multiplexer based on coupled line defect channels

The objective of this paper is to investigate the implementation of a hybrid photonic crystal (PhC) 1.31/1.55 μm wavelength division multiplexer (WDM) and wavelength channel interleaver with channel spacing of roughly 0.8 nm between the operating wavelengths of 1.54-1.56 μm. It is based on 1-D photonic crystal (PhC) structure connected with an output 2-D PhC structure. The power transfer efficiency of the hybrid PhC WDM at 1.31 μm and 1.55 μm were computed by eigen-mode expansion (EME) method to be about 88% at both the wavelengths. The extinction ratios obtained for the 1.31 μm and 1.55 μm wavelengths are − 25.8 dB and − 22.9 dB respectively.     

The thermo-optical variable optical attenuator based on SOI

A thermo-optical variable optical attenuator was studied based on silicon on insulator (SOI) substrate waveguide. It is composed by the single-mode waveguide, taper waveguide, multi-mode waveguide, and inclined electrode. By adjusting the applied voltage on the inclined electrode it can achieve continuously variable attenuation of the output light. The results we studied show that when the applied voltage is about 4.7 V (the corresponding power is 233 mW), the variation of the waveguide's core temperature is about 33 °C, the refractive index changes more than 5.0 × 10⁻³ and the attenuation will reach 35 dB, and the response time is only 35 μs.     

FWM in silicon nanocrystal-based sandwiched slot waveguides

The observation of four-wave mixing in a 6 mm long sandwiched slot waveguide filled with Si-nc/SiO₂ is reported for optical powers usually employed in telecommunication systems. A −47 dB conversion efficiency is measured in fabricated waveguides for input signal powers around 12 dBm on chip and a waveguide length of 6 mm. Furthermore, the calculated non-linear coefficient is found to be n₂ = 2.67 × 10⁻¹⁷ m²/W. It is also expected that, by using longer waveguides, it would be possible to achieve wavelength conversion.     

Micromachined high frequency PMN-PT/epoxy 1-3 composite ultrasonic annular array

This paper reports the design, fabrication, and performance of miniature micromachined high frequency PMN-PT/epoxy 1-3 composite ultrasonic annular arrays. The PMN-PT single crystal 1-3 composites were made with micromachining techniques. The area of a single crystal pillar was 9 × 9 μm. The width of the kerf among pillars was ∼5 μm and the kerfs were filled with a polymer. The composite thickness was 25 μm. A six-element annular transducer of equal element area of 0.2 mm² with 16 μm kerf widths between annuli was produced. The aperture size the array transducer is about 1.5 mm in diameter. A novel electrical interconnection strategy for high density array elements was implemented. After the transducer was attached to the electric connection board and packaged, the array transducer was tested in a pulse/echo arrangement, whereby the center frequency, bandwidth, two-way insertion loss (IL), and cross talk between adjacent elements were measured for each annulus. The center frequency was 50 MHz and −6 dB bandwidth was 90%. The average insertion loss was 19.5 dB at 50 MHz and the crosstalk between adjacent elements was about −35 dB. The micromachining techniques described in this paper are promising for the fabrication of other types of high frequency transducers, e.g. 1D and 2D arrays.