Color demosaicing using variance of color differences.

This paper presents an adaptive demosaicing algorithm. Missing green samples are first estimated based on the variances of the color differences along different edge directions. The missing red and blue components are then estimated based on the interpolated green plane. This algorithm can effectively preserve the details in texture regions and, at the same time, it can significantly reduce the color artifacts. As compared with the latest demosaicing algorithms, the proposed algorithm produces the best average demosaicing performance both objectively and subjectively.     

Current gain dependence on subcollector and etch-stop doping inInGaP/GaAs HBTs

The DC current gain dependence of InGaP/GaAs heterojunction bipolar transistors (HBTs) on subcollector and etch-stop doping is examined. Samples of InGaP/GaAs HBTs having various combinations of subcollector doping and etch-stop doping are grown, and large area 60 μm×60 (μ) HBTs are then fabricated for DC characterization. It is found that the DC current gain has a strong dependence on the doping concentration in the subcollector and the subcollector etch-stop. Maximum gain is achieved when the subcollector is doped at 6~7×10 ¹⁸ cm^-3 while the subcollector etch-stop is doped either above 6×10¹⁸ cm^-3 (current gain/sheet resistance ratio, β/R_b=0.435 at I_c=1 mA) or below 3.5×10¹⁷ cm^-3 (β/R_b=0.426~0.438 at I_c=1 mA). The data show that it is not necessary to heavily dope the subcollector etch-stop to reduce the conduction barrier and to obtain high current gain. The high current gain obtained with the low InGaP etch-stop doping concentration is attributed to the reduction of the effective energy barrier thickness due to band bending at the heterojunction between the InGaP etch-stop and the GaAs subcollector. These results show that the β/R_b of InGaP/GaAs HBTs can improve as much as 69% with the optimized doping concentration in subcollector and subcollector etch-stop     

A self-driven active clamp forward converter using the auxiliary winding of the power transformer

This study proposes a new self-driven active clamp forward converter eliminating the extra drive circuit for the active clamp switch. The converter used the auxiliary winding of the power transformer to drive the active clamp switch and a simple RC circuit to get the dead time between the two switches. The operation principle was presented and experimental results were used to verify theoretical predictions. A 100-W (5 V/20 A) prototype converter that only exhibited 1.5-turn winding number in the auxiliary winding was sufficient to drive the active clamp switch on the input of 50 V. Finally, the measured efficiency of the converter was presented and the maximum efficiency of 91% was obtained.     

Experimental Demonstration of 1.25-Gb/s Radio-Over-Fiber Downlink Using SBS-Based Photonic Upconversion

We experimentally demonstrate a radio-over-fiber downlink system using a stimulated Brillouin scattering (SBS)-based photonic upconversion technique. The Brillouin selective amplification characteristic of SBS is incorporated to generate the 11-GHz band radio-frequency (RF) carrier. The dual-electrode Mach-Zehnder optical modulator, which is used to carry the broadband data in the optical carrier instead of the optical sideband, is adopted along with the SBS-based carrier generation setup. To vindicate the broadband capabilities of the proposed scheme, 1.25-Gb/s pseudorandom bit sequence data is carried in the optical carrier. Error-free operation of the 1.25-Gb/s downlink is achieved without critical power penalties after the 13-km fiber transmission.     

An analytical threshold-voltage model of trench-isolated MOSdevices with nonuniformly doped substrates

A simple closed-form expression of the threshold voltage is developed for trench-isolated MOS (TIMOS) devices with feature size down to the deep-submicrometer range. The analytical expression is the first developed to include the nonuniform doping effect of a narrow-gate-width device. The inverse narrow width effect can be predicted analytically from the proposed model. It was derived by modeling the gate sidewall capacitance to include the two-dimensional field-induced edge fringing effect and solving the Poisson equation to include the channel implant effect at different operating backgate biases. A two-dimensional simulation program was developed, and the simulated data were used for verification of the analytical model. Good agreements between the modeled and simulated data have been achieved for a wide range of gate widths and biases. The model is well suited for the design of the basic transistor cell in DRAM circuits using trench field oxide isolation structure     

Proton-exchange optical waveguides in Z-cut LiNbO/sub 3/ using toluic acid

Proton-exchanged planar waveguides were demonstrated in Z-cut LiNBO/sub 3/ using toluic acid as a new organic proton source. These waveguides exhibit a propagation loss of around 1 dB/cm, and a step refractive index profile with an index increase of 0.124 measured at 0.663 mu m. The diffusion rate was found to be lower than those obtained using the popular benzoic and phosphoric acids.<>     

Reliability analysis of a human operator under different levels of stress

This paper presents a reliability analysis of a human operator carrying out his jobs under different levels of stress. The failed system, which is due to self-correctable error, is restored to as-new at normal stress. Laplace transforms of state probabilities and operator reliability are derived. The mean time to human error is also given.     

Isometry-Based Shape-Adaptive Fractal Coding for Images

Fractal image coding is an effective method to eliminate the image redundancy through piecewise self-transformability. The fractal code consists of a set of contractive affine transforms. To improve the performance when a range block experiences large error, we usually partition the range block into square or nonsquare subrange blocks for two- or multilevel fractal coding. In this paper, we find an inherent property of fractal coding that can be used to decide the edge orientation of a range block. Then this property is used for shape-adaptive fractal coding (SAFC). In SAFC, the top-level range block is partitioned into square or nonsquare (rectangle or triangle) subrange blocks for multilevel fractal encoding. Here, the maximum size of the range block can be the same as that of the whole image size while the minimum size is 4×4. In SAFC, no additional computations are required to obtain the edge orientation of a range block. Instead, we propose an edge-orientation detector, where the edge orientation of a range block is obtained during the fractal encoding process. According to our simulation results, SAFC can reduce the bit rate requirement of the conventional fractal coding scheme.     

Coherent frequency/phase modulation scheme

An L-phase-levelled coherent frequency/phase modulation (LCFPM) scheme is proposed and analysed for additive white Gaussian noise channels. The optimised LCFPM scheme with L>2 is found to have superior bandwidth efficiency to the conventional dual-frequency quadrature frequency/phase modulation scheme