Previous Projects - 1997

Renee Ball - Ohio University

Project: Revamp & Interface, Advisor: Prof. S.Wagner

This project consists of taking apart a machine that measures the conductivity of an amorphous silicon sample with respect to temperature. The machine's parts were replaced and resoderd. The next step is to interface the hardware with software on a PC, labview. The hardware end of the project is complete, but the software is taking longer than expected. I do not foresee finishing the software end of the project but I should have the program started.

Project: Alignment of Textual Images, Advisor: Prof. P. Ramadge

The purpose of this project was to study the problem of automatically aligning images which share common regions. In this study, we focus on textual images, which vary by translation and/or rotation. While brute force search is prohibitively costly in terms of time, alignment using random selection of small "patches" has been demonstrated to be effective.

Project: Watermarking for Digital Images, Advisor: Prof. S. Schwartz

In this project we seek to study the effectiveness of a more robust detection process using simple non-parametric detectors such as the sign detector. The effectiveness of the detectors are tested by various common non malicious image manipulations processes. Some examples of these processes include the following: additive gaussian noise, additive salt & pepper noise, bilinear rotation, blurring, JPEG compression, color quantization (from more to lesser entry colormaps). The probability of detection failure and false detection are used to determine the effectiveness of the detectors.

Project: Maximum - Likelihood Receivers for the GSM Digital Cellular Standard, Advisor: Prof. H. Kobayashi

This project involves working on developing improved receivers for the GSM pan-European digital cellular system. Though the transmitted signal cannot be modified (it is specified by and must adhere to the GSM standard), the manufacturer of mobile phones and base statrons is free to use any technique to receive the data. Along with other members of the professors group, I have been working on an iterative decoding scheme using ambiguity zone detection.

Project: SPAM Compiler, Advisor: Prof. S. Malik

This project involved saving and restoring the data structures of the back end of the Spam compiler. The saving past involved writing the data structures to a file. The restoring part involved reading the file and building the structures again. The outcome of the project could be used for testing the various optimizations of the compiler. I also debugged parts of the compiler. I was successful in detecting a few bugs, which were then taken care of.

Project: Performance Measurement Statistics, Advisor: Prof. M. Martonosi

This Project involved design and construction of a tool for evaluating software performance based on hardware statistics stored in machine specific registers on the Pentium and Pentium Pro Processors. In this project a Kernel-level driver and a virtual level driver was written that were controlled by the use r-level interface to gain access to the Pentium registers.

Project: Low Noise Pre-Amplifier for Measuring Johnson noise at Cryogenic Temperatures, Advisor: Prof. D. Tsui

My project was to design, build, and test a low noise voltage pre-amplifier for use at cryogenic temperatures. The primary function of such an amplifier is to amplify weak Johnson noise (thermal noise) signals (1nV or less) from a two dimensional electron gas at a temperature of 100mK. From the Johnson noise we can then determine the temperature of the electrons given the following relationship: en^2=4*kb*R*T*B, where en= Johnson noise, kb=Boltzman constant, T=temperature (K), and B=band width. Once the temperate of the electrons can be determined we can determine whether or not the electrons are at the ambient temperature of the cryogenic environment.

Project: Converting the Source Code of a Rolodex Program, Advisor: Prof. J. Plett

This project involved converting the source code of a Rolodex program for one graphical package to another. The student first needed to become familiar with the package that he/she was converting and then proceed to fix the quirks in the code that confounded the automatic conversion program. The source code was converted to run in the new graphical environment and then cleaned up so that the different functions were grouped together by purpose and were clearly commented.

Project: Work Function Changes in Organic Semiconductors, Advisor: Prof. A. Kahn

This project consisted of measuring semiconductor work function changes under various conditions, using the Kelvin Probe. Silicon, Gallium Arsenide and Gold samples were used to verify the accuracy and repeatability of the measurements since their work function changes were known from previous experiments. However, repeatable data was never achieved and retooling/troubleshooting the equipment has been an ongoing task. The ultimate goal of this project was to measure the work function changes, especially surface photovoltage, of organic semiconductors.

Project: Non-Diffracting Planar Beams in Unbiased Photorefractive Media, Advisor: Prof. M. Segev

In starting this project the first experimental observation of a novel type of non-diffracting beam was recorded. A hologram was used to create an Airy function from Gaussian beam. The self-focusing properties of the photorefractive crystal compensate the diffraction of the Airy beam, causing the beam to propagate without diffraction in the crystal.

Project: Heat Transfer in OLED Display, Advisor: Prof. J. Sturm

This project was to determine the heat transfer coefficient as a function of characteristic size Lc. (Defined as area/perimeter of the plate). This is a concern because the OLED's break down a temp. which is too high. A fair amount of power is needed to run such a display however, so the device tends to heat up a quite a bit. So this is a critical issue in the commercial development of OLED Displays.

Project: Module Addressing Feature for the SPAM Compiler, Advisor: Prof. S. Malik

This project was to debug the SPAM compiler and implement module addressing capability for Motorola 56K DSP chips. In debugging we found a couple of existing bugs in the compiler. The module addressing capability should greatly speed up input of a stream of data.