|
5 |
|
|
Description The objective of the proposed research is developing a quantitative estimation of the reliability of software system. The Monte Carlo method is a very useful tool when system complexity makes the formulation of exact models impossible. This characteristic of the Monte Carlo method makes it an ideal tool for estimating the reliability of software-based system. Unlike many other mathematical models that used for software reliability predication, the system complexity is irrelevant to the method. The method can predict the software reliability even in the lack of the precise information on the structure of system software components and the system dynamic. Instead, system components need only be tested for failure during operation, which ensure that components, which are used more often, contribute proportionally more to the overall reliability estimate. Combined with self-checking algorithms, which respond to randomly generated inputs, the method obviates the need for valid, nontrivial input data and external oracle. |
|
|
6 |
|
|
Description Lead poisoning can cause many serious diseases in human beings who are exposed to mine areas. Diseases like lung cancer, asthma, allergies, have been reported from Mine areas in Colorado. The focus of this research is to create an Information base on Genomics from a specific mining area from State of Colorado. Visualization and cross comparison of this information will be helpful in understanding the relation between the degree of exposure and the extent of damage at genetic level. This would provide a better treatment for such diseases. We feel that this effort is important because there is no computer based system which analyzes the damage due to lead poisoning from a variety of mines in Colorado at this time. Our research will primarily focus on: 1. Biochemical isolation and sequencing of genetic material (DNA and Protein) from the cells exposed to higher concentration of lead compounds. 2. This sequence would be analyzed with the help of NCBI/BLAST software to see the degree of damage by comparing those with healthy cells. 3. Analysis of 2D/3D data on Genetic material and creation of Gene data bank for quick analysis and access to be developed at the University of Colorado, Colorado Springs using OpenGL and C, or Java3D. We are working on gathering samples of related patients data or the sequence data from health science center and National Jewish Medical r esearch center in Denver. The focus of our work would be on the efficient visualization and comparison of these data sets.
|
|
|
7 |
|
|
Description It is proposed to develop an encoding/decoding algorithm using overcomplete wavelets. They will be used as building blocks for construction of communication signals, which represent binary data vectors. This approach should lead to a maximized throughput (number of users) as well as to the robust to noise recovery of user data. In addition, the algorithm is expected to ensure secure communication or low probability of the intercept of user data. The aforementioned characteristics of the wavelet-based encoding/decoding algorithm were confirmed by a preliminary feasibility study. Once this algorithm is developed, its performance will be evaluated using proper quantitative tools, which are to be developed in the project. |
|
|
8 |
|
|
Description Signal processing is a very important component of communication. Often there is a need to do some kind of signal estimation, system identification, spectral analysis, inverse channel identification, noise cancellation or prediction. Many of these problems would occur in communications signal processing or in the tracking of a vehicle or target. I am proposing to do research to solve a problem in one of the above areas using perhaps an FIR or IIR filtering or maybe an adaptive filtering algorithms such as LMS for stationary processes or Kalman filtering in the case of a non-stationary process such as the tracking of a target as it changes position, etc. |
|