H optimal approximation for causal spline interpolation

In this paper, we give a causal solution to the problem of spline interpolation using H^∞ optimal approximation. Generally speaking, spline interpolation requires filtering the whole sampled data, the past and the future, to reconstruct the inter-sample values. This leads to non-causality of the filter, and this becomes a critical issue for real-time applications.Our objective here is to derive a causal system which approximates spline interpolation by H^∞ optimization for the filter. The advantage of H^∞ optimization is that it can address uncertainty in the input signals to be interpolated in design, and hence the optimized system has robustness property against signal uncertainty.We give a closed-form solution to the H^∞ optimization in the case of the cubic splines. For higher-order splines, the optimal filter can be effectively solved by a numerical computation. We also show that the optimal FIR (finite impulse response) filter can be designed by an LMI (linear matrix inequality), which can also be effectively solved numerically. A design example is presented to illustrate the result.     

Improvement of output voltage control performance for low-speed operation of matrix converter

The matrix converter (MxC) is an ac-to-ac conversion device that can generate variable frequency and variable voltage output. The nine bidirectional switches of an MxC allow pulse-width modulation (PWM) control of input currents and output voltages. PWM switches need "switch commutations" from one switch to another. During these switch commutations, however, unwanted voltage error occurs like a dead time effect in a voltage source inverter (VSI). When PWM pulse-widths are narrower than the time for completing a commutation sequence, voltage error rapidly increases. In the low-speed range, PWM pulses become narrower and voltage distortions become larger due to incomplete commutations. Moreover, these errors are critical in the low-speed operation because the system is sensitive to the smallest voltage error. In this paper, a new PWM strategy is proposed for improving voltage control performance in the low-speed range. Based on the input and output voltage information, PWM pulse-widths are controlled to avoid incomplete commutation. The feasibility of the proposed method is proved through simulation and experimental results.     

Upconversion pumped thulium-doped fluoride fiber amplifier andlaser operating at 1.47 μm

A 1.064-μm band upconversion pumped Tm³⁺-doped fluoride fiber amplifier and a laser both operating at 1.47 μm are investigated in detail. The two devices are based on the ³F ₄→³H₄ transition in a trivalent thulium ion, which is a self-terminating system. When pumped at 1.064 μm, the amplifier has a gain of over 10 dB from 1.44 to 1.51 μm and a low-noise characteristic. Also, the fiber laser generates a high-output power of over 100 mW with a slope efficiency of 59% at around 1.47 μm. These levels of performance will be important for optical communication systems     

Observation of iron silicide formation by plan-view transmission electron microscopy

The formation and the phase transitions of iron silicide by solid-phase epitaxy have been investigated by means of plan-view transmission electron microscopy, which enables us to observe a clean interface between Fe and Si. Layers of Fe were deposited on Si (100) at room temperature in an ultrahigh vacuum chamber. The sample was annealed in the electron microscope at a temperature between 673 and 1073 K. After annealing at 673 K, FeSi crystallites were formed with various orientations. When the annealing temperature was increased to 973 K, we found that the crystallites suddenly started to coalesce into grains of several hundreds of nanometers in size and polycrystalline beta-FeSi2 was formed. These phase transitions were also confirmed with electron energy-loss spectroscopy.     

HDR satellite codec with concatenated multilevel codes

The performance of a high data rate (HDR) codec based on a multidimensional modulation code derived from multilevel codes is evaluated for a satellite channel. The codec reliably supports HDR transmission for broadband ISDN over a 72 MHz satellite transponder and is designed to be used with an 8 PSK modem. Results show that significant coding gain with higher spectral efficiency can be achieved compared to uncoded QPSK     

NRZ operation at 40 Gb/s of a compact module containing an MQW electroabsorption modulator integrated with a DFB laser

NRZ operation at 40 Gb/s has been successfully performed using a very compact module of a multiple-quantum-well (MQW) electroabsorption modulator integrated with a distributed-feedback (DFB) laser. While the DFB laser is injected with a constant current, the integrated MQW electroabsorption modulator is driven with a 10-Gb/s electrical NRZ signal. A clearly opened eye diagram has been observed in the modulated light from the modulator. And a receiver sensitivity of -27.2 dBm at 10/sup -9/ has been experimentally confirmed in the bit-error-rate (BER) performance.     

Design criteria of 1.3-μm multiple-quantum-well lasers forhigh-temperature operation

We present design criteria for high-temperature operation in 1.3-μm multiple-quantum-well (MQW) lasers from the viewpoint of the light output power penalty, i.e., the change in the light output power at a fixed drive current with increasing temperature. It is shown that not only the characteristic temperature (T₀) but also internal loss dependence on temperature (γ) and threshold current (I_th) are significant parameters for reducing the power penalty. We compare the high-temperature performance of InGaAsP-based and AlGaInAs-based MQW lasers and demonstrate that AlGaInAs-based lasers have more potential in terms of the power penalty. Furthermore, we also demonstrate that the power penalty can be reduced by introducing a buried-heterostructure (BH) structure into AlGaInAs-based lasers. From these results, we conclude that the AlGaInAs-based BH lasers are promising for high-temperature performance     

Fabrication of MOD-derived YBCO films on [001]LaAlO/sub 3/ and their application to /spl lambda//4 CPW SIR BPFs

Chemically derived epitaxial thin films of YBa/sub 2/Cu/sub 3/O/sub 7-/spl delta// (YBCO) are fabricated on [001]LaAlO/sub 3/ substrates by the metalorganic-deposition (MOD) process, which has advantages of high quality, nonvacuum, low-cost, and large-scale production of high-T/sub c/ superconducting films. The MOD-derived YBCO films have a sharp transition at the critical temperature (90.4 K) and a high-quality film with a surface resistance of 0.13 m/spl Omega/ (30 K, 9.98 GHz) is obtained. As a microwave application, simple and compact bandpass filters (BPFs) using /spl lambda//4 coplanar-waveguide. stepped-impedance resonators are demonstrated on the YBCO films. A two-stage Chebyshev BPF of center frequency of 5.731 GHz, bandwidth of 135 MHz, and insertion loss of 0.29 dB with little input power dependency in a power range less than 10 dBm is realized on the film.     

Electron transport behaviour in Nb-doped SrTiO3 bicrystals

In order to elucidate the relationship between electron transport behaviour and defect chemistry, grain boundary structure and current-voltage characteristics across the boundary were investigated for Nb-doped SrTiO3 bicrystals. Two kinds of boundaries, i.e. small angle and random boundaries, were prepared for 0.2at% and 1.0at% Nb-doped SrTiO3. The bicrystals were prepared by joining two single crystals at 1400 degrees C for 10 h under a pressure of 0.4 MPa in air. High-resolution transmission electron microscopy (HRTEM) study revealed that all of the joined boundaries are free from any secondary phases or amorphous films. On the other hand, it was found that non-linearity in current-voltage dependence becomes remarkable by reduction of cooling rate after joining in small angle boundaries of 0.2at% Nb-doped SrTiO3 bicrystal. In addition, the random boundary of 1.0at% Nb-doped SrTiO3 bicrystal exhibits clear alpha = 2 I-V relation, which appears across a contact of semiconductor-insulator-semiconductor (n-i-n). From the results of HRTEM study and I-V behaviours, it could be concluded that the electron transport mechanism is controlled mainly by defect chemistry and not by the grain boundary structure.     

Generation of picosecond blue light pulse by frequency doubling ofgain-switched GaAlAs laser diode having saturable absorber

Picosecond pulse generation of blue light by frequency doubling of a GaAlAs laser diode is reported. High power pulse generation is realized by incorporating gain switching of a laser diode with a saturable absorber and frequency doubling in a proton-exchanged MgO:LiNbO₃ waveguide. The laser diode with a longer saturable absorber can produce optical pulses with higher peak power and narrower pulse width. The spectral bandwidth of second-harmonic generation for the waveguide is evaluated at about 20 nm. This is wide enough to frequency-double all the multilongitudinal modes of the gain-switched laser diode. A blue light pulse of 7.88-mW maximum peak power and 28.7-ps pulsewidth is obtained for a 1.23-W peak pulse of the laser diode