Summer Research Experience for Undergraduates (REU)
 

Home 
Philosophy 
Application 
Schedule 
Past Projects 
Facilities 
UM-TC Area 
Housing 
For Teachers 
Mentors 

1999 Summer Abstracts

 

Carrie Wilson - Amalgam of Software Technologies for the Elimination of Expensive Inefficiencies in Medical Information Flow and Handling
 
Judy Linkhart - Aluminum Deposition Al/GaInP Schottky Diodes
 
Michelle Prampin - Comparison of Average Output SNR of MRC and EGC Receivers Over Nakagami Fading Channels
 
Samuel Jiawei Ng - Thermal Expansion Coeffieients of Interference Coating Films for Optical Mems
 
Timothy Holston - Using the Java Speculative MultiThreading (JavaSpMT) for Parallel Processing
 
Matthew Rundquist - Three Dimensional Modeling a Tuned TEM Resonator for Computational Electromagnetic Simulation and Testing
 
Rodney Dent - Screensaver FDTD Applications of UNIX and Java
 
Kyle Willis - Modeling Radio Channel Multipath Fading with Nakagami and Rice Parameters
 
Elisha Cohen - Magnetic Recording Heads
 
Kelly Seabold - Java Speculative MultiThreading
 
Angela Bartow - The InTENsity PowerWall
 
Lakisha N. Patterson - GIS: Geographic Information Systems A Java Interface
 
Margaret R. Horton - High Moment Magnetic Materials For Inductive Writing Head Applications
 
Peter Schommer - Er3+ and Pr3+ Doped Waveguide Amplifiers
 
Aaron Walters - Continuous Level Sensing for Flexible Venous Reservoir

1999 Summer Program Staff

bar03_dot3x3_darkyellow_2.gif

 

 

                               

Professor Douglas Ernie           Lisa Youcha                 Professor Lori Lucke

Program Assistant

 

 

 

1999 Summer Program Abstracts

bar03_dot3x3_darkyellow_3.gif

Participant: Carrie Wilson
Home Institution:
University of Toledo
Faculty Mentor:
Dr. James Holte
Project:
Amalgamation of Software Technologies for the Elimination of Expensive Inefficiencies in Medical Information Flow and Handling

Whenever any type of medical transaction occurs, it is necessary to record the event into a patient file; this exam information is used for insurance and billing as well as medical history purposes. Our research in the area of medical records is concerned with the path traversal from the acquisition of the medical details from the source to the final destination, the record, and specifically direct solutions for the elimination of gross inefficiencies in the traversal of patient information from the point of the visit with the examining physician through the supporting structure of the medical institution. Process such as transcription of medical dictation and deciphering of medical claims codes, tasks which currently consume expensive manual labor and thusly transfer into health care cost as an expensive factor, can be automated with integration and development of the right technology. Transcription automation is achieved by implementation of window software which operates in cooperation with speech recognition to serve as a template for the collection and handling of patient exam information. Claims coding is automated with an innovative efficiency which interprets exam information by directing dictation into key component fields of the application which are interpreted by the implemented coding algorithm, bystepping the need for complex extraction of key coding components from already transcribed data. By developing applications which directly function as specific improvements to transcription and code inefficiencies, we have synthesized an inexpensive, straightforward solution and thus facilitated an immediate improvement in information flow by providing a tool which yields significant improvement in time and cost efficiency within the current working system.

Participant: Judy Linkhart
Home Institution:
Seattle Pacific University
Faculty Mentor:
Dr. Marshall Nathan
Project: Aluminum Deposition Al/GaInP Schottky Diodes

N-type Al/GaInP Schottky diodes were grown using Molecular Beam Epitaxy and Electron Beam Evaporation. Both samples had GaInP grown on to p of a GaAs substrate using Molecular Beam Epitaxal growth. The Schottky barrier heights were measured using C-V and I-V curves. The diode ideality factor was also determined from I-V curves. The material with Electron Beam deposited Aluminum had poor Schottky contacts demonstrated by barrier heights that were 0.33eV different for I-V and C-V measurements. The material with MBE deposited Aluminum had credible barrier heights that were 0.08eV different for C-V and I-V measurements. The barrier height determined by O-V measurements was 0.57eV and the barrier height determined by C-V measurements was 0.65eV. Molecular Beam Epitaxal growth was shown to be the preferred method of Aluminum deposition for Al/GaInP/GaAs Schottky barrier diodes.

Participant: Michelle Prampin
Home Institution:
Xavier University of LA
Faculty Mentor:
Professor Mohamed Slim Alouini
Project: Comparison of Average Output SNR of MRC and EGC Receivers Over Nakagami Fading Channels

 
A cellular telephone system provides a wireless link to the public switched telephone system (PSTN). Antennas located at base stations, or in cell phones, receive multiple copies of a certain transmitted frequency. Each signal is out of phase and varies in its signal to noise ratio (SNR), or signal intensity. A diversity combiner, Maximal Ratio Combiner (MRC) or Equal Gain Combiner(EGC), combines all frequencies together into one signal that maximizes its signal power. This project presents a theoretical and simulated analysis of these two combiners. A theoretical analysis requires deriving closed form exact and asymptotic equations of the two combiners. Simulated results verifies that the theoretical outcome is correct. Although MRC give optimum performance among all linear diversity combiners, using EGC at times saves power, costs, and complexity.

Participant: Samuel Jiawei Ng
Home Institution:
University of California, Irvine
Faculty Mentor:
Dr. Joseph J. Talghader
Project: Thermal Expansion Coefficients of Interference Coating Films for Optical Mems

This paper investigates one of the mechanical properties of interference coating films: the coefficient of thermal expansion (CTE). Knowledge of this property is crucial in designing thermally stable and interferometrically flat micromirrors, or micromirrors that do not curve for a given change in temperature. This is desirable in applications such as magneto-optical recording and barcode scanning where an optical beam is focused on a particular area. The CTE is found by studying the change in a bilayer cantilever's radius of curvature with varying temperatures. This depends on the cantilever's deflection due to interference of coherent light, calculated by observing a series of interference fringes.

Participant: Timothy Holston
Home Institution:
Mississippi Valley State University
Faculty Mentor:
Dr. David Lilja
Graduate Student Mentor: Iffat Kazi
Project: Using the JAVA Speculative Multithreading (JavaSpMT) for Parallel Processing

I started this project with intentions to decrease processing time of programs that are know for there long processing time. To do this, a serial program is reviewed and it is evaluated to see if the program can be parallellized to make the processing time faster. In doing this, you would have to manual look at the program and go through and evaluate a loop to see if that particular loop is where time is consumed.

 

Participant: Matthew Rundquist
Home Institution:
St. John's University
Faculty Mentor:
Dr. Matthew O'Keefe
Project: Three Dimensional Modeling a Tuned TEM Resonator for Computational Electromagnetic

A computer-simulated, three-dimensional tuned transverse electromagnetic (TEM) resonator has been developed, which incorporates a human head that has fifteen characterized tissue-types. The resonator is based off J.T. Vaughan's tuned TEM resonators that are being built and tested in the Center for Magnetic Resonance (CMRR), at the University of Minnesota. The simulation has a resolution sized of one mm and is the physical TEM resonator and to learn more about the interaction between the electromagnetic waves of the resonator and the human head.

Participant: Rodney Dent
Home Institution:
Mississippi Valley State University
Faculty Mentor: Dr.
Matthew O'Keefe
Project: Screensaver FDTD Applications of UNIX and JAVA

Screensaver FDTD was designed to enhance my knowledge of computer operating system and programming languages. The operating system that was introduced was the UNIX operating system. The use of the commands and shells provide the ability for many things to be accomplished. The programming language was Java. Before venturing into Screensaver FDTD, I was only familiar with the Windows operating systems. Creating Screensaver FDTD required me to learn how to use UNIX. This project also gave me a good introduction to Java. While venturing in the world of UNIX, I learned about it's portability. The portability of UNIX allows it to be used on many makes of computers. Portability comes in very handy in programming. While learning both UNIX and Java, the concepts of computing and programming that I had learned at Mississippi Valley State University were also needed to be successful. As a result of touching basis with my prior education in the field of Computer Science, I added on to those things as well.

Participant: Kyle Willis
Home Institution:
University of South Florida
Faculty Mentor:
Professor Mohamed Slim Alouini
Graduate Student Mentor: Ali Abdi
Project: Modeling Radio Channel Multipath Fading with Nakagami and Rice Parameters

Data is collected from a mobile receiver representing a real radio signal envelop. Distributions of the multipath (fast) fading component of the various radio channels are presented. The Nakagami and Rice first order Cumulative Distribution Functions (CDF) are fitted to the empirical curves of these fast fading signals. Kolmogorov-Smirnoff (maximum error) statistics are generated for each distribution. The second order statistics, Level Crossing Rate (LCR) and Average Fade Duration (AFD), are also fit to empirical curves. The mean Square Error (MSE) is calculated for each data file. Overall analysis of the accuracy of the models based on visual observations and goodness of fit statistics is presented for both the Nakagami and Rice models.

Participant: Elisha Cohen
Home Institution:
University of Wisconsin Madison
Project: Magnetic Recording Heads

Participant: Kelly Seabold
Home Institution:
Xavier University
Faculty Mentor: Dr.
David Lilja
Graduate Student Mentor: Iffat Kazi
Project: JAVA Speculative Multithreading

Java is an extremely portable language, however, as an interpreted language, this portability also limits Java programs to slower execution that compiled languages, i.e. C, C++. Serial Java programs with implicit parallelism can be optimized using language level support to run on multiple threads. The Speculative Multi-Threading model used for this project utilizes control speculation and runtime analysis of dependent data. This gives us the ability to work with any data structure, a more powerful execution model than other previous parallelization models for Java. By converting serial programs using this model, Java is a viable alternative to compiled languages with comparable execution time an preserved portability. This model was evaluated by converting a program for the SPEC JVM98 benchmark suite, compress.

Participant: Angela Bartow
Home Institution:
Kansas State University
Faculty Mentor:
Dr. David Du
Project: The InTENsityTM PowerWallTM

Researchers at the University of Minnesota's Laboratory for Computational Science and Engineering have created a semi-immersive environment which fills a user's field of view at a resolution of 6400X2048 pixels. The InTENsityTM PowerWallTM uses rear-projection technology to provide high resolution imaging of scientific data in collaborative or presentation environment. This paper presents a brief history of the InTENsityTM PowerWallTM and its predecessor. The current system configuration is then described in terms of the individual components used to construct the entire system. More details is then given regarding the numerous networks that join all the components together into a working system. I sate my conclusions to date about this project and describe the future work that needs to be done. Also included in appendices are network design diagrams, an equipment list, equipment product guides and datasheets, and pictures.

Participant: Lakisha N. Patterson
Home Institution:
Langston University
Faculty Mentor:
Professor Shashi Shekhar
Project: GIS: Geographic Information Systems, a JAVA Interface

Geographic Information Systems (GIS) is a high-tech equivalent of a map. In other words it does the exact same things a map does, except it does more. For example, a map is use to find out how to get from place to place while traveling, but it is impossible to use a map to answer spatial queries. That is where GIS comes in. This system allows a person to find out the characteristics of a place such as, how cold or how it is, that or the crime rate in that area. GIS also answers spatial queries such as, the country that has the maximum number of neighbors of the highest point in Canada. Since GIS cannot read regular map information, the information has to be put into digital form. This is where the Database Management System (DBMS) comes in. The DBMS is a system used for storing, manipulating, and retrieving data for the GIS. Once all information has been gathered, it can be put in the DBMS and stored for later use.

Participant: Margaret R. Horton
Home Institution:
Trinity University
Faculty Mentor:
Professor Philip I. Cohen
Project: High Moment Magnetic Materials for Inductive Writing Head Applications

The focus of the research is to create a material for use in the inductive recording head of computer hard drives with high saturation magnetization, low remanent magnetization and a uniaxial anisotropy sufficient to maintain physical orientation during the writing process. Such a material would be able to write on the necessary highly coercive recording media of next-generation hard drives. The currently-used material, Permalloy (Fe79Ni21), has a soon-approaching limit in the maximum coercivity of media onto which it can write. iron nitride materials are an alternative solution to Permalloy except that FeN lacks stability and uniaxial anisotropy. By creating microscopic corrugations photolithographically impressed onto a Si substrate and depositing FeNiTaN, it is believed that a stable FeN phase with a uniaxial anisotropy can be created by pair-ordering of the Ni atoms adsorbed onto the atomic steps of the corrugations. Towards this goal, Si wafers with corrugations 1um deep and 2um wide spaced 1 um apart were patterned and Rf sputtered with and Fe target in an Ar and N2 mixture. X-ray diffraction and Auger electron spectroscopy were used to find the composition of the films. The Scanning Electron Microscope allowed the topology of the patterned sputtered wafers to be observed and verified. While iron nitride was indeed deposited, throughout the twelve completed sputtering runs with varied process parameters each time, a strong iron oxide presence remained.

Participant: Peter Schommer
Home Institution:
St. John's University
Faculty Mentor:
Dr. Anand Gopinath
Project: Er3+ an Pr3+ Doped Waveguide Amplifiers

Dielectric planar waveguide amplifiers will be important to communications devices in the future. By doping wafers with different Rare Earth elements, one can amplify signals that are used in communications. Originally, Xe sputtered wafers were made by doping Al2O3 with Er and by doping ZrO2 with Pr/Yb. Recently, Ar sputtered wafers have been acquired and limited testing has been done on those. life times increased from 3.8ms in the Xe samples to 5.3ms in the Ar samples for 0.5% Er doping concentrations. Losses for Xe sputtered Er samples were 4.31 dB/cm at 1.0A and 3.80 dB/cm at 2.0A for the 980nm pump. The 1530nm signal had 5.43 dB/cm loss at 100 mA. Ar sputtered wafers are currently being tested.

Participant: Aaron Walters
Home Institution:
University of Wyoming
Faculty Mentor:
Dr. Lori Lucke
Project: Continuous Level Sensing for Flexible Venous Reservoir

This project focuses on a very specific part of the extremely complicated process of cardiopulmonary bypass surgery. During cardiopulmonary bypass surgery the blood must be oxygenated and circulated artificially outside of the body. As seen in Figure 1, this circulation is accomplished with a pump, arterial and venous lines, and a blood reservoir. First oxygenated blood flows out to the body through the arterial lines. The body tissues use the oxygen from the blood then return deoxygenated through the venous lines. The venous return is once again oxygenated and the cycle starts again. The flow of blood is driven with a peristaltic roller pump of centrifugal pump. The blood reservoir ensures that increasing and decreasing demands of blood due to the stresses of surgery will be met.
     For the external circulation system to be successful, it is critical to prevent air bubbles from entering the external flow of blood. One area where the circuit is vulnerable to the introduction of an air bubble is the blood reservoir. The reservoirs used in the system are currently of two types, a hard plastic reservoir, and a flexible (soft) reservoir.
The hard plastic reservoir is more apt to introduce air bubbles into the system. As the fluid level decreases more air surrounds the blood increasing the chance of introduction of an air bubble. However, the hard shell container allows for simple electronic level detection. Acoustic sensors are placed at various levels on the container. When these sensors are calibrated for the difference between air and liquid the level of blood in the reservoir can be automatically detected.
     A soft shell reservoir is less likely to introduce air bubbles to the system. The container collapses and expands with the liquid to prevent air from entering the system. however, because the container readily changes shape, acoustic level detection becomes difficulty. Currently, a special attendant, called a Perfusionist, must physically watch the level of blood in the reservoir and make adjustments to the flow in the circuit by hand. This is a boring repetitive process. If electronic level sensing were available, the flow could be adjusted automatically reducing the chance of human error. Automatic level detection for soft shell reservoirs is the focus of this project.