Wavelet-based low-delay ECG compression algorithm for continuous ECG transmission.

The delay performance of compression algorithms is particularly important when time-critical data transmission is required. In this paper, we propose a wavelet-based electrocardiogram (ECG) compression algorithm with a low delay property for instantaneous, continuous ECG transmission suitable for telecardiology applications over a wireless network. The proposed algorithm reduces the frame size as much as possible to achieve a low delay, while maintaining reconstructed signal quality. To attain both low delay and high quality, it employs waveform partitioning, adaptive frame size adjustment, wavelet compression, flexible bit allocation, and header compression. The performances of the proposed algorithm in terms of reconstructed signal quality, processing delay, and error resilience were evaluated using the Massachusetts Institute of Technology University and Beth Israel Hospital (MIT-BIH) and Creighton University Ventricular Tachyarrhythmia (CU) databases and a code division multiple access-based simulation model with mobile channel noise.     

Growth of InGaN HBTs by MOCVD

The design and growth of GaN/InGaN heterojunction bipolar transistors (HBTs) by metalorganic chemical vapor deposition (MOCVD) are studied. Atomic-force microscopy (AFM) images of p⁺InGaN base layers (∼100 nm) deposited under various growth conditions indicate that the optimal growth temperature is limited to the range between 810 and 830°C due to a trade-off between surface roughness and indium incorporation. At these temperatures, the growth pressure must be kept above 300 Torr in order to keep surface pit density under control. An InGaN graded-composition emitter is adopted in order to reduce the number of V-shaped defects, which appear at the interface between GaN emitter and InGaN base and render an abrupt emitter-base heterojunction nearly impossible. However, the device performance is severely limited by the high p-type base contact resistance due to surface etching damage, which resulted from the emitter mesa etch.     

Effect of ferrosilicon addition on the composition of inclusions in 16Cr-14Ni-Si stainless steel melts

The silicon deoxidation equilibrium between the 16Cr-14Ni-1.5Mn-Si melts and the CaO-SiO₂-8MgO-5CaF₂ (basicity=1.8) slag at 1743 K was investigated to understand the effect of aluminum and silicon contents on the composition of inclusions. Therefore, the ferrosilicon alloys with different aluminum content were chosen based on the preceding objective. In addition, the phase stability diagram of the inclusions was computed using commercial thermodynamic software based on the Gibbs energy minimization principles. The content of MnO in the inclusions sharply decreases with increasing silicon content when the steel melts were deoxidized by the ferrosilicon alloys containing high aluminum (FeSi-H). The content of SiO₂ in the inclusions slightly increases with increasing silicon content when the FeSi-L is used, while a maximum value is shown at [Si]=1.5 pct when the FeSi-H is used. The content of MgO in the inclusions increases by increasing the content of silicon, regardless of the kinds of ferrosilicon alloys. The use of the FeSi-L as a deoxidizer could suppress the formation of Al₂O₃ in the inclusions, while the content of Al₂O₃ increases with increasing silicon content when the FeSi-H is used. When the FeSi-H is used as a deoxidizer, the inclusions are the glassy type with the composition of Mn-silicates at [Si]≤1.3 pct, while the Mg(Ca)-silicates with the composition of the forsterite phase are observed in the steel composition of [Si]=3.3 pct. When the steel melts were deoxidized by the FeSi-L alloys, the inclusions are the glassy-type Mn-silicates at [Si]=0.8 pct, while the Mn-silicates containing the cristobalite phase are observed at [Si]=1.5 to 2.4 pct. In the composition of [Si]=3.3 pct, the Mg-silicates with the composition of the rhodonite phase are observed. The log(X _SiO2/X _MnO) of the inclusions linearly increases by increasing the log [a _Si · a _O / a _Mn] with the slope close to unity when the FeSi-L is used as a deoxidizer, while the slope of the line is about 2 times greater than that of the expected value when FeSi-H is used. The log (X _MgO/X _MnO) of the inclusions linearly increases by increasing the log [a _Mg/a _Mn] with slopes greater than the expected value of unity.     

Novel Approach to Integration of Numerical Modeling and Field Observations for Deep Excavations

Precedent and observation of performance are an essential part of the design and construction process in geotechnical engineering. For deep urban excavations designers rely on empirical data to estimate potential deformations and impact on surrounding structures. Numerical simulations are also employed to estimate induced ground deformations. Significant resources are dedicated to monitor construction activities and control induced ground deformations. While engineers are able to learn from observations, numerical simulations have been unable to fully benefit from information gained at a given site or prior excavation case histories in the same area. A novel analysis method, self-learning in engineering simulations (SelfSim), is introduced to integrate precedent into numerical simulations. SelfSim is an inverse analysis technique that combines finite element method, biologically inspired material models, and field measurements. SelfSim extracts relevant constitutive soil information from field measurements of excavation response such as lateral wall deformations and surface settlement. The resulting soil model, used in a numerical analysis, provides correct ground deformations and can be used in estimating deformations of similar excavations. The soil model can continuously evolve using additional field information. SelfSim is demonstrated using two excavation case histories in Boston and Chicago.     

High fill-factor two-axis gimbaled tip-tilt-piston micromirror array actuated by self-Aligned vertical electrostatic combdrives

In this paper, we present a high fill-factor micromirror array actuated by self-aligned vertical electrostatic combdrives. To meet the requirements of applications in free-space communication and imaging, each micromirror has three degrees of freedom of motion: rotation around two axes in the mirror plane and linear translation perpendicular to the mirror plane. Our approach is to integrate the high fill-factor reflectors into the fabrication process of the actuators on the wafer-scale. Multilevel silicon-on-insulator (SOI) bonding is utilized to form the high optical quality reflectors and high aspect-ratio vertical combdrive actuators. The wiring for electrical access to the multielectrode per pixel array is fabricated on separate wafers by thin film processing, and flip-chip bonded to the reflector/actuator chip. This architecture overcomes the fill-factor limitation of top-side accessed electrical addressing of mirrors made on SOI. Our 360/spl mu/m pixel size mirror array achieves a 99% fill-factor with optically flat reflectors.     

A scalable hybrid approach for extracting head components from Web tables

We have established a preprocessing method for determining the meaningfulness of a table to allow for information extraction from tables on the Internet. A table offers a preeminent clue in text mining because it contains meaningful data displayed in rows and columns. However, tables are used on the Internet for both knowledge structuring and document design. Therefore, we were interested in determining whether or not a table has meaningfulness that is related to the structural information provided at the abstraction level of the table head. Accordingly, we: 1) investigated the types of tables present in HTML documents, 2) established the features that distinguished meaningful tables from others, 3) constructed a training data set using the established features after having filtered any obvious decorative tables, and 4) constructed a classification model using a decision tree. Based on these features, we set up heuristics for table head extraction from meaningful tables, and obtained an F-measure of 95.0 percent in distinguishing meaningful tables from decorative tables and an accuracy of 82.1 percent in extracting the table head from the meaningful tables.     

Solvability of Centralized Supervisory Control Under Partial Observation

The problem of synthesizing a nontrivial controllable and observable sublanguage of a given non-prefix-closed language is addressed. This problem arises in supervisory control of discrete-event systems, when the objective is to synthesize safe nonblocking supervisors for partially observed systems. The decentralized version of this problem is known to be unsolvable. We show that the centralized version of this problem is solvable by presenting a new algorithm that synthesizes a nontrivial controllable and observable sublanguage of the given non-prefix-closed language, if one exists. We also show that the union of all nonblocking solutions to the associated supervisory control problem can be expressed as the union of all regular nonblocking solutions. This work was done when the first author was at the University of Michigan as a Ph.D. student.     

谁该为信用证欺诈买单？

由于外贸体制和法律的不健全,我国成为贸易大国的同时,也是信用证欺诈的受害大国自我国加入世贸组织以来,国内市场逐步对外开放。随着2006年7月1日最新修订的《对外贸易法》实施,将有更多的企业、经济组织和个人加入到外贸行列中来,我国的国际贸易市场将越来越活跃。在国际贸易     

A new modeling interface for the pen-input displays

Sketch interactions based on interpreting multiple pen markings into a 3D shape is easy to design but not to use. First of all, it is difficult for the user to memorize a complete set of pen markings for a certain 3D shape. Secondly, the system will be waiting for the user to complete the sequence of the pen markings, often causing a certain mode error. To address these problems, we present a novel, interaction framework, suitable for interpretations based on single-stroke marking on pen-input display; within this framework 3D shape modeling operations are designed to create appropriate communication protocols.     

Intelligent System for Adapting to a User’s Characteristics

In this paper, we propose an intelligent system that adapts itself to a user’s characteristics or habits. The proposed intelligent system is composed of two Learning Vector Quantisation (LVQ) networks, commonly used in the fields of pattern recognition and signal processing. From the external condition of the plant, the first LVQ network learns to recognise the pattern of the sensed signal, and then aids the second LVQ to learn the user’s characteristics or habits so as to automatically produce the user’s favoured output. To verify the usefulness of the proposed method, we simulated and experimented with a variable illuminator. Both simulation and experimental results showed that the proposed intelligent system learns to automatically produce the illuminator output that the user most favours for the circumstances.