# Mathematics, computer science and statistics.

Chair: Andrew Harrell, CEWES-GM

Vice-chair: Elgenaid Hamadain, Jackson State University

THURSDAY MORNING

Meeting Room 1

9:00 MODIFIED ALPHABET OVERLAP GRAPHS

Veranda Moffett, Mississippi Valley State University, Itta Bena, MS 38941

A graph is a collection of points and lines connecting some subset of these lines or points. The points of a graph are most commonly known as graph vertices, but may also be called nodes or points. Similarly, the lines connecting the vertices of a graph are most commonly known as graph edges, but may also be called arcs or lines. The purpose of this paper is to give an example of how to construct a regular graph from an Alphabet Overlap graph. Throughout this research I studied Alphabet Overlap graphs where the k*n vertices are labeled with the sequences of length n from an alphabet of size k. Two vertices u and v are joined by an edge if and only if the first w digits of u are identical to the last w digits of v. The Alphabet Overlap graph that I studied is denoted by AO (2,k, k-1) where a- is the size of the alphabet, k- the length of the sequence, and t- the length of the overlap. In this presentation I will show under what conditions is it possible to modify a size two Alphabet Overlap graph, so that it is a regular graph of degree four. I will also compare what is known about large cubic graphs and compare any possible results of the cubic graphs to the modified Alphabet Overlap graph.

9:30 ALGEBRAIC PROPERTIES OF SUM ONE MATRICES

Lenton McLendon* and Joseph Kolibal, University of Southern Mississippi, Hattiesburg, MS 39406

The Sum One Matrices (SOM), i.e., those whose columns or rows sum to one, have useful algebraic properties. This class subsumes the stochastic matrices, which are of interest in constructing stochastic interpolation and approximation methods. We examine some of these properties in relation to the problem of stochastic interpolation, with particular interest in improving the numerical efficiency of these methods.

10:00 ESTIMATE AND HANDLE DATA ERRORS IN GEOGRAPHIC INFORMATION SYSTEM: A LITERATURE STUDY

Nicholas Black and Lixin Yu*, Alcorn State University, Lorman, MS 39096

The precision and accuracy of spatial data are important factors that may influence the reliability of the Geographic Information Systems output. This research project surveyed the related documents of the past ten years to identify the issues that have been studied to improve the precision of the spatial data and to reduce the data errors. It used several case studies to demonstrate the importance of being able to estimate the impact of the data errors. It also surveyed the techniques used to reduce data errors in the GIS applications. The literature study shows how GIS, GPS, and digital elevation model technologies are used together to make measurement. In some cases, digital calculation is used in replacement of the actual measurement in order to reduce the demand of resource, knowing that this practice could introduce some random errors. This can be done as long as the impact of the errors is correctly estimated and the degree of errors is within an allowable range.

10:30 INTERPOLATION SCHEMES FOR SENSOR DATA FROM SURFACES WITH FRACTAL TEXTURE

Andrew W. Harrell, Engineering Research and Development Center, Vicksburg, MS 39180

This talk with discuss various ways to interpolate data from information on sensor surfaces using 1 and 2 dimensional fractal generation programs. MATLAB programs were written to plot dimensional Brownian motion surfaces from their Hurst exponents using inverse Fourier transforms. Histograms of the power spectrum of the original data and the data from the interpolated surface were compared. The effects of using different types of Gaussian functions and white noise function in the subroutine that generates the Brownian motion data were investigated. Programs in MATLAB to do mid-point interpolation algorithms were written and the results compared, in terms of the histograms of the power spectrums, with the inverse fourier transform approach. Also, programs that use multi-fractals to do the interpolation were written and the effect of the lacunarity parameter of the fractal investigated as to the goodness of fit of the interpolated surface with the orginal sensor data.

11:00 WEIGH IN MOTION (WIM) INTERFACING WITH TC-AIMS II AND AALPS

Sabrina Phillips, Mississippi Valley State University, Itta Bena, MS 38941

The Department of Energy, Oak Ridge Operations Office (DOE-ORO) oversees advanced research and development programs through its management and operating contractors. One of those contractors, UT-Battelle, LLC, manages and operates the Oak Ridge National Laboratory (ORNL) for DOE-ORO. ORNL has highly specialized, and often unique, scientific, and information technology capabilities available to solve nations level problems including critical national technology challenges. ORNL's highly specialized multi-disciplinary and comprehensive approaches offer solutions that are not available elsewhere in the public private sectors. Additionally, ORNL approaches are unbiased and independent of commercial considerations. Currently the Army manually identifies the vehicles and enters this information into the joint Transportation Coordinators' Automated Information for Movement System II (TC-AIMS II). The Army also weighs vehicles, manually calculates vehicle individual axle weights, total vehicular weight and manually measures the length of the vehicle. They then calculate the center of balance data and manually mark this information on the vehicle. This information is then manually transferred to the Automated Air Load Planning System (AALPS) personnel who manually enter it into the AALPS system. Each of these steps in the process is prone to human error. By establishing (1) an automated data exchange between the vehicle Radio Frequency Identification (RFID) tags and WIM; and (2) an automated data exchange link between WIM and TC-AIMS II/AALPS, identification planning data can be automatically transferred from TC-AIMS II to WIM and "actual" weight data from WIM to AALPS thus eliminating those human errors and at the same time expediting the process. Critical to the establishment of this data exchange is the modeling, design, and implementation efforts that are documenting the data/information process flow through the system. Key Use Cases, Activity Diagrams, Sequence Diagrams and Collaboration Diagrams capturing the data/information flows are presented herein

11:30 REGULARLY STRUCTURED SUM ONE MATRICES

Lenton McLendon, University of Southern Mississippi, Hattiesburg, MS 39406

In this research we examine the properties of classical examples of full matrices, concentrating on sum one matrices, attempting to develop more fully the properties of row stochastic matrices with cyclic row symmetries. This pattern in the coefficients of a matrix is typical of the algebraic structure which arises from working with the discretization of symmetric integral operators, such as those associated with discrete deconvolution.

THURSDAY AFTERNOON

Meeting Room 1

2:00 Divisional Business Meeting

3:00 Divisional Poster Session--Exhibit Hall B

ROBOTIC FORMATION: CORRECTING BOE-BOT'S ERROR BY TRIAL INSTALLATION

Tisha Brown, Mississippi Valley State University, Itta Bena, MS 38941

Military strategist use robotic formation as a defense tactic in many battle field maneuvers. Consequently, the superior government funded technology and machines used in their research are a far cry from resources of the average researcher. This research will use a Stamp Module microcontroller to perform specific platform activities and logical formation. The programming language used to obtain feedback is similar to the BASIC software. The software editor provides a step-by-step reference and seamless installation programs. Subsystem testing is essential to the detection of errors prior to construction of Board of Education, chassis and module. The brain of the operation is a module that use artificial intelligence in an attempt to accurately and successfully program the controlled machines at optimum performance levels. However, before this can be achieved research must be implemented to test the light, touch, sensory and navigational skills of the robot in a real world environment. A proposed method to detect moderate to maximum behavioral commands is the assumption of programming the Stamp Module to execute exercises using straight line, triangular, figure-eight, photo resistors and infrared interference: Lead-Shadow accuracy Boe-Bot testing. Therefore, an attempt to view all aspects of the features included on the Boe-Bot is analyzed, critiqued and recorded.

A METHOD TO PROPERLY COLOR AN ALPHABET OVERLAP GRAPH

Glenda Span, Mississippi Valley State University, Itta Bena, MS 38941

We define a graph G to be an alphabet overlap graph denoted by G=AO(a,k,t), where the letters a, k, and t represent the size of the alphabet, length of the sequence, and length of the overlap and where AO stands for "alphabet overlap. The a*k vertices are each labeled with one of the sequences of length k from an alphabet of size a. Two vertices are adjacent if the corresponding tags are the same. While studying the particular graph G=AO(2, k, k-1), methods for finding the chromatic number were explored. In this work, we give an algorithm for properly coloring the Alphabet overlap graphs G=AO(2, k, k-1).

AUTOMATED SOFTWARE TESTING USING RATIONAL ROBOT

Cedric Foster, Mississippi Valley State University, Itta Bena, MS 38941

Automated testing is the use of strategies, tools and artifacts that reduce the need for manual and human involvement when trying to perform a specific task. Rational Robot is a licensed automated testing tool that is developed by IBM. It is a functional and performance test tool for software developing teams that want to automate regression testing. The research conducted this summer consisted of testing NASA's RAMS SQL software using Rational Robot. RAMS SQL is the software application that documents the authorization and performance of work that is done on the Space Shuttle Main Engine. The general purpose of this research was to execute functional and regression testing on the RAMS SQL software to test how new revisions applied to the application interacted with its pre-existing functionalities. These functional tests were carried out using test cases which gave instructions on how to test the software by executing different keystrokes and mouse clicks. Rational Test Manager played a huge part in during this research also. Rational Test Manager allowed me to do distributed functional testing, run suites, and view test logs immediately after testing was completed to see whether the RAMS SQL application failed or passed the testing it was under. While experimenting with Robot, I learned how to install license keys, create projects, run test scripts in series, and access projects from different computers.

FRIDAY MORNING

Exhibit Hall A2

Special Subsession on Supercomputing and Concurrent Poster Session

8:30 CARR-PARINELLO MOLECULAR DYNAMICS BENCHMARK PERFORMANCE

Alexandr Isayev and Taner Pirim*, Jackson State University, Jackson, MS 39217 and Mississippi Center for Supercomputing Research, University, MS 38677

Carr-Parinello Molecular Dynamics (CPMD) is a well-parallelized, plane wave/pseudopotential software implementation of Density Functional Theory. CPMD has been certified to run on many architectures including 32-bit Intel clusters of PCs, SGI's Altix 3000 family of global shared-memory, 64-bit Itanium2, and high performance compute servers. However, there are many variances in how the source code may be installed on a given architecture, and configuration information on successful Altix installations is not yet readily available. The potential performance of a particular CPMD problem on a given system is influenced not only by the hardware architecture of the system itself, but also on such variables as the brand and version of compiler used in the installation, and the implementation and version of third-party math libraries called by the CPMD code. At the Mississippi Center for Supercomputing Research (MCSR), it is important that chemistry researchers run their calculations using the application, system, and processor/disk/CPU settings that will maximize their efficiency and throughput. In this study, CPMD's own benchmarks are used to investigate the relative performance of CPMD installation on MCSR's SGI Altix 3700 global shared-memory system, and comparison of the results obtained has been made to the results of various UNIX platform high performance computers such as CRAY as well as Beowulf Linux Cluster.

8:55 IMPLEMENTING A TABU SEARCH ALGORITHM FOR THE CONTACT MAP ALIGNMENT PROBLEM USING OPENMP AND MPI

Jason Hale* and Wei Liu, University of Mississippi, University, MS 38677

An unconstrained binary quadratic programming (UBQP) model has been proposed by Liu (2005) to solve the contact map alignment problem in protein structure comparison. This research explores parallel tabu search algorithms for the contact map alignment problem, and for UBQP problems in general. The algorithms are coded in C++, and executed on a shared-memory supercomputer, and a distributed memory computer cluster, at the Mississippi Center for Supercomputing Research. Results are empirically evaluated.

9:20 AN ALGORITHM TO DETERMINE THE SEQUENCE OF STABLE MULTICAST TREES IN MOBILE AD HOC NETWORKS

Natarajan Meghanathan, Jackson State University, Jackson, MS 39217

Given the source node, the set of receivers of a multicast session, and the knowledge of future topology changes, we propose an optimal polynomial-time algorithm called OptTreeTrans to determine the minimum number of multicast tree transitions during a multicast session in mobile ad hoc networks (MANETs). Algorithm OptTreeTrans operates based on the following greedy heuristic: Whenever a multicast tree is required to connect a source to all its receivers, choose the tree that will exist for the longest time. The above strategy is repeated over the duration of the multicast session. A sequence of such stable multicast trees is called the stable mobile multicast tree (SMMT). Though there are only heuristics to approximate the minimum number of links in a multicast tree, we prove that algorithm OptTreeTrans gives the optimal number of tree transitions, and simultaneously yields the SMMT. We also introduce the notion of "look-ahead window size", as the time for which information about future topology changes are known. We study the performance of OptTreeTrans in terms of the number of tree transitions and tree size (i.e., number of links constituting the multicast tree) for different values of look-ahead window size, node mobility, network density and multicast group size. Results indicate that the stability of multicast trees in MANETs could be improved significantly by looking at the near future. For a given node mobility, we also observe a tradeoff between number of tree transitions and tree size in terms of look-ahead window size, network density and multicast group size.

9:40 MISSISSIPPI CENTER FOR SUPERCOMPUTING RESEARCH (MCSR) USER ADVISORY GROUP MEETING, POSTER SESSION, AND SPECIAL SUBSESSION ON SUPERCOMPUTING

David G. Roach, University of Mississippi, University, MS 38677

The Mississippi Center for Supercomputing Research was established in 1987 by the Mississippi Legislature and the Institutions of Higher Learning (IHL) in order to provide high performance supercomputing (HPC) support for research and instruction at all state universities. The Mississippi Supercomputer User Advisory Committee (MSUAG) was established by the IHL Research Consortium to provide user input and advice to MCSR management and technical staff on policies and procedures for the Center's operations. It includes member representatives from all IHL institutions. The Advisory Group will meet at this MAS conference. Mr. David G. Roach, Director of the MCSR, will conduct the meeting. The agenda includes an update on MCSR HPC facilities and services, introduction of new MCSR staff members, and site reports and ongoing research updates by MSUAG representatives. A poster session, showcasing research projects that utilize MCSR facilities and services, will follow the Advisory Group Meeting. A Special HPC Subsession of the Mathematics, Computer Science, and Statistics Division, sponsored by the MCSR, will also be held to serve as a forum on supercomputing in which faculty and graduate student researchers will have the opportunity to describe their research projects that involve HPC, Internet2, Grid Computing, Visualization, Network Security, Computer Systems Administration, and the use of MCSR resources. IHL faculty and graduate students, with an interest in HPC and/or MCSR facilities and services, are also invited to attend and participate.

10:30 Break

Regular Session Resumes

10:45 COMPARISON ANALYSIS OF HUMAN PROTEIN DATABASES AND APPROACH TO DATABASE INTEROPERABILITY

Arvinder Kang, University of Mississippi, University, MS 38677

There are growing number of different proteomic databases available. However the data and curation in each database differs to a large extent. This research discusses three different such databases and the difference in their XML structure. The scope of interoperability is discussed and Human Proteome Organization's Proteomics Standards Initiative is looked at as a promising approach. Biobuilder, a database development and functional annotation platform, powering Human Protein Reference Database, is an implementation of these new standards favoring synchronization and interoperability.

FRIDAY AFTERNOON

Exhibit Hall A2

12:30 Voting on Awards for Best Presentation and Poster and Presentation of Awards

1:00 MATHEMATICS, COMPUTER, AND INFORMATION SCIENCES ALUMNI SURVEY ONLINE FORM

John Stepney, Mississippi Valley State University, Itta Bena, MS 38941

In this project an online MCIS alumni survey form is created that connects to an Oracle database and automatically creates an online report. The MCIS alumni survey allows the alumni of Mississippi Valley State University's Mathematics, Computer, and Information Sciences department alumni to complete the survey and submit their data online. The database is created using Oracle database management system and the Structure Query Language. After the information enters the Oracle database then an online report is automatically generated based on the data from the database. The purpose of this project is to allow the MCIS department to generate summative data about the department and its graduates. The MCIS online survey will make retrieving the alumni's information faster and more efficient. The MCIS alumni survey will be transferred to the Mississippi Valley State University server to allow MCIS alumni to access the form off campus.

Vice-chair: Elgenaid Hamadain, Jackson State University

THURSDAY MORNING

Meeting Room 1

9:00 MODIFIED ALPHABET OVERLAP GRAPHS

Veranda Moffett, Mississippi Valley State University, Itta Bena, MS 38941

A graph is a collection of points and lines connecting some subset of these lines or points. The points of a graph are most commonly known as graph vertices, but may also be called nodes or points. Similarly, the lines connecting the vertices of a graph are most commonly known as graph edges, but may also be called arcs or lines. The purpose of this paper is to give an example of how to construct a regular graph from an Alphabet Overlap graph. Throughout this research I studied Alphabet Overlap graphs where the k*n vertices are labeled with the sequences of length n from an alphabet of size k. Two vertices u and v are joined by an edge if and only if the first w digits of u are identical to the last w digits of v. The Alphabet Overlap graph that I studied is denoted by AO (2,k, k-1) where a- is the size of the alphabet, k- the length of the sequence, and t- the length of the overlap. In this presentation I will show under what conditions is it possible to modify a size two Alphabet Overlap graph, so that it is a regular graph of degree four. I will also compare what is known about large cubic graphs and compare any possible results of the cubic graphs to the modified Alphabet Overlap graph.

9:30 ALGEBRAIC PROPERTIES OF SUM ONE MATRICES

Lenton McLendon* and Joseph Kolibal, University of Southern Mississippi, Hattiesburg, MS 39406

The Sum One Matrices (SOM), i.e., those whose columns or rows sum to one, have useful algebraic properties. This class subsumes the stochastic matrices, which are of interest in constructing stochastic interpolation and approximation methods. We examine some of these properties in relation to the problem of stochastic interpolation, with particular interest in improving the numerical efficiency of these methods.

10:00 ESTIMATE AND HANDLE DATA ERRORS IN GEOGRAPHIC INFORMATION SYSTEM: A LITERATURE STUDY

Nicholas Black and Lixin Yu*, Alcorn State University, Lorman, MS 39096

The precision and accuracy of spatial data are important factors that may influence the reliability of the Geographic Information Systems output. This research project surveyed the related documents of the past ten years to identify the issues that have been studied to improve the precision of the spatial data and to reduce the data errors. It used several case studies to demonstrate the importance of being able to estimate the impact of the data errors. It also surveyed the techniques used to reduce data errors in the GIS applications. The literature study shows how GIS, GPS, and digital elevation model technologies are used together to make measurement. In some cases, digital calculation is used in replacement of the actual measurement in order to reduce the demand of resource, knowing that this practice could introduce some random errors. This can be done as long as the impact of the errors is correctly estimated and the degree of errors is within an allowable range.

10:30 INTERPOLATION SCHEMES FOR SENSOR DATA FROM SURFACES WITH FRACTAL TEXTURE

Andrew W. Harrell, Engineering Research and Development Center, Vicksburg, MS 39180

This talk with discuss various ways to interpolate data from information on sensor surfaces using 1 and 2 dimensional fractal generation programs. MATLAB programs were written to plot dimensional Brownian motion surfaces from their Hurst exponents using inverse Fourier transforms. Histograms of the power spectrum of the original data and the data from the interpolated surface were compared. The effects of using different types of Gaussian functions and white noise function in the subroutine that generates the Brownian motion data were investigated. Programs in MATLAB to do mid-point interpolation algorithms were written and the results compared, in terms of the histograms of the power spectrums, with the inverse fourier transform approach. Also, programs that use multi-fractals to do the interpolation were written and the effect of the lacunarity parameter of the fractal investigated as to the goodness of fit of the interpolated surface with the orginal sensor data.

11:00 WEIGH IN MOTION (WIM) INTERFACING WITH TC-AIMS II AND AALPS

Sabrina Phillips, Mississippi Valley State University, Itta Bena, MS 38941

The Department of Energy, Oak Ridge Operations Office (DOE-ORO) oversees advanced research and development programs through its management and operating contractors. One of those contractors, UT-Battelle, LLC, manages and operates the Oak Ridge National Laboratory (ORNL) for DOE-ORO. ORNL has highly specialized, and often unique, scientific, and information technology capabilities available to solve nations level problems including critical national technology challenges. ORNL's highly specialized multi-disciplinary and comprehensive approaches offer solutions that are not available elsewhere in the public private sectors. Additionally, ORNL approaches are unbiased and independent of commercial considerations. Currently the Army manually identifies the vehicles and enters this information into the joint Transportation Coordinators' Automated Information for Movement System II (TC-AIMS II). The Army also weighs vehicles, manually calculates vehicle individual axle weights, total vehicular weight and manually measures the length of the vehicle. They then calculate the center of balance data and manually mark this information on the vehicle. This information is then manually transferred to the Automated Air Load Planning System (AALPS) personnel who manually enter it into the AALPS system. Each of these steps in the process is prone to human error. By establishing (1) an automated data exchange between the vehicle Radio Frequency Identification (RFID) tags and WIM; and (2) an automated data exchange link between WIM and TC-AIMS II/AALPS, identification planning data can be automatically transferred from TC-AIMS II to WIM and "actual" weight data from WIM to AALPS thus eliminating those human errors and at the same time expediting the process. Critical to the establishment of this data exchange is the modeling, design, and implementation efforts that are documenting the data/information process flow through the system. Key Use Cases, Activity Diagrams, Sequence Diagrams and Collaboration Diagrams capturing the data/information flows are presented herein

11:30 REGULARLY STRUCTURED SUM ONE MATRICES

Lenton McLendon, University of Southern Mississippi, Hattiesburg, MS 39406

In this research we examine the properties of classical examples of full matrices, concentrating on sum one matrices, attempting to develop more fully the properties of row stochastic matrices with cyclic row symmetries. This pattern in the coefficients of a matrix is typical of the algebraic structure which arises from working with the discretization of symmetric integral operators, such as those associated with discrete deconvolution.

THURSDAY AFTERNOON

Meeting Room 1

2:00 Divisional Business Meeting

3:00 Divisional Poster Session--Exhibit Hall B

ROBOTIC FORMATION: CORRECTING BOE-BOT'S ERROR BY TRIAL INSTALLATION

Tisha Brown, Mississippi Valley State University, Itta Bena, MS 38941

Military strategist use robotic formation as a defense tactic in many battle field maneuvers. Consequently, the superior government funded technology and machines used in their research are a far cry from resources of the average researcher. This research will use a Stamp Module microcontroller to perform specific platform activities and logical formation. The programming language used to obtain feedback is similar to the BASIC software. The software editor provides a step-by-step reference and seamless installation programs. Subsystem testing is essential to the detection of errors prior to construction of Board of Education, chassis and module. The brain of the operation is a module that use artificial intelligence in an attempt to accurately and successfully program the controlled machines at optimum performance levels. However, before this can be achieved research must be implemented to test the light, touch, sensory and navigational skills of the robot in a real world environment. A proposed method to detect moderate to maximum behavioral commands is the assumption of programming the Stamp Module to execute exercises using straight line, triangular, figure-eight, photo resistors and infrared interference: Lead-Shadow accuracy Boe-Bot testing. Therefore, an attempt to view all aspects of the features included on the Boe-Bot is analyzed, critiqued and recorded.

A METHOD TO PROPERLY COLOR AN ALPHABET OVERLAP GRAPH

Glenda Span, Mississippi Valley State University, Itta Bena, MS 38941

We define a graph G to be an alphabet overlap graph denoted by G=AO(a,k,t), where the letters a, k, and t represent the size of the alphabet, length of the sequence, and length of the overlap and where AO stands for "alphabet overlap. The a*k vertices are each labeled with one of the sequences of length k from an alphabet of size a. Two vertices are adjacent if the corresponding tags are the same. While studying the particular graph G=AO(2, k, k-1), methods for finding the chromatic number were explored. In this work, we give an algorithm for properly coloring the Alphabet overlap graphs G=AO(2, k, k-1).

AUTOMATED SOFTWARE TESTING USING RATIONAL ROBOT

Cedric Foster, Mississippi Valley State University, Itta Bena, MS 38941

Automated testing is the use of strategies, tools and artifacts that reduce the need for manual and human involvement when trying to perform a specific task. Rational Robot is a licensed automated testing tool that is developed by IBM. It is a functional and performance test tool for software developing teams that want to automate regression testing. The research conducted this summer consisted of testing NASA's RAMS SQL software using Rational Robot. RAMS SQL is the software application that documents the authorization and performance of work that is done on the Space Shuttle Main Engine. The general purpose of this research was to execute functional and regression testing on the RAMS SQL software to test how new revisions applied to the application interacted with its pre-existing functionalities. These functional tests were carried out using test cases which gave instructions on how to test the software by executing different keystrokes and mouse clicks. Rational Test Manager played a huge part in during this research also. Rational Test Manager allowed me to do distributed functional testing, run suites, and view test logs immediately after testing was completed to see whether the RAMS SQL application failed or passed the testing it was under. While experimenting with Robot, I learned how to install license keys, create projects, run test scripts in series, and access projects from different computers.

FRIDAY MORNING

Exhibit Hall A2

Special Subsession on Supercomputing and Concurrent Poster Session

8:30 CARR-PARINELLO MOLECULAR DYNAMICS BENCHMARK PERFORMANCE

Alexandr Isayev and Taner Pirim*, Jackson State University, Jackson, MS 39217 and Mississippi Center for Supercomputing Research, University, MS 38677

Carr-Parinello Molecular Dynamics (CPMD) is a well-parallelized, plane wave/pseudopotential software implementation of Density Functional Theory. CPMD has been certified to run on many architectures including 32-bit Intel clusters of PCs, SGI's Altix 3000 family of global shared-memory, 64-bit Itanium2, and high performance compute servers. However, there are many variances in how the source code may be installed on a given architecture, and configuration information on successful Altix installations is not yet readily available. The potential performance of a particular CPMD problem on a given system is influenced not only by the hardware architecture of the system itself, but also on such variables as the brand and version of compiler used in the installation, and the implementation and version of third-party math libraries called by the CPMD code. At the Mississippi Center for Supercomputing Research (MCSR), it is important that chemistry researchers run their calculations using the application, system, and processor/disk/CPU settings that will maximize their efficiency and throughput. In this study, CPMD's own benchmarks are used to investigate the relative performance of CPMD installation on MCSR's SGI Altix 3700 global shared-memory system, and comparison of the results obtained has been made to the results of various UNIX platform high performance computers such as CRAY as well as Beowulf Linux Cluster.

8:55 IMPLEMENTING A TABU SEARCH ALGORITHM FOR THE CONTACT MAP ALIGNMENT PROBLEM USING OPENMP AND MPI

Jason Hale* and Wei Liu, University of Mississippi, University, MS 38677

An unconstrained binary quadratic programming (UBQP) model has been proposed by Liu (2005) to solve the contact map alignment problem in protein structure comparison. This research explores parallel tabu search algorithms for the contact map alignment problem, and for UBQP problems in general. The algorithms are coded in C++, and executed on a shared-memory supercomputer, and a distributed memory computer cluster, at the Mississippi Center for Supercomputing Research. Results are empirically evaluated.

9:20 AN ALGORITHM TO DETERMINE THE SEQUENCE OF STABLE MULTICAST TREES IN MOBILE AD HOC NETWORKS

Natarajan Meghanathan, Jackson State University, Jackson, MS 39217

Given the source node, the set of receivers of a multicast session, and the knowledge of future topology changes, we propose an optimal polynomial-time algorithm called OptTreeTrans to determine the minimum number of multicast tree transitions during a multicast session in mobile ad hoc networks (MANETs). Algorithm OptTreeTrans operates based on the following greedy heuristic: Whenever a multicast tree is required to connect a source to all its receivers, choose the tree that will exist for the longest time. The above strategy is repeated over the duration of the multicast session. A sequence of such stable multicast trees is called the stable mobile multicast tree (SMMT). Though there are only heuristics to approximate the minimum number of links in a multicast tree, we prove that algorithm OptTreeTrans gives the optimal number of tree transitions, and simultaneously yields the SMMT. We also introduce the notion of "look-ahead window size", as the time for which information about future topology changes are known. We study the performance of OptTreeTrans in terms of the number of tree transitions and tree size (i.e., number of links constituting the multicast tree) for different values of look-ahead window size, node mobility, network density and multicast group size. Results indicate that the stability of multicast trees in MANETs could be improved significantly by looking at the near future. For a given node mobility, we also observe a tradeoff between number of tree transitions and tree size in terms of look-ahead window size, network density and multicast group size.

9:40 MISSISSIPPI CENTER FOR SUPERCOMPUTING RESEARCH (MCSR) USER ADVISORY GROUP MEETING, POSTER SESSION, AND SPECIAL SUBSESSION ON SUPERCOMPUTING

David G. Roach, University of Mississippi, University, MS 38677

The Mississippi Center for Supercomputing Research was established in 1987 by the Mississippi Legislature and the Institutions of Higher Learning (IHL) in order to provide high performance supercomputing (HPC) support for research and instruction at all state universities. The Mississippi Supercomputer User Advisory Committee (MSUAG) was established by the IHL Research Consortium to provide user input and advice to MCSR management and technical staff on policies and procedures for the Center's operations. It includes member representatives from all IHL institutions. The Advisory Group will meet at this MAS conference. Mr. David G. Roach, Director of the MCSR, will conduct the meeting. The agenda includes an update on MCSR HPC facilities and services, introduction of new MCSR staff members, and site reports and ongoing research updates by MSUAG representatives. A poster session, showcasing research projects that utilize MCSR facilities and services, will follow the Advisory Group Meeting. A Special HPC Subsession of the Mathematics, Computer Science, and Statistics Division, sponsored by the MCSR, will also be held to serve as a forum on supercomputing in which faculty and graduate student researchers will have the opportunity to describe their research projects that involve HPC, Internet2, Grid Computing, Visualization, Network Security, Computer Systems Administration, and the use of MCSR resources. IHL faculty and graduate students, with an interest in HPC and/or MCSR facilities and services, are also invited to attend and participate.

10:30 Break

Regular Session Resumes

10:45 COMPARISON ANALYSIS OF HUMAN PROTEIN DATABASES AND APPROACH TO DATABASE INTEROPERABILITY

Arvinder Kang, University of Mississippi, University, MS 38677

There are growing number of different proteomic databases available. However the data and curation in each database differs to a large extent. This research discusses three different such databases and the difference in their XML structure. The scope of interoperability is discussed and Human Proteome Organization's Proteomics Standards Initiative is looked at as a promising approach. Biobuilder, a database development and functional annotation platform, powering Human Protein Reference Database, is an implementation of these new standards favoring synchronization and interoperability.

FRIDAY AFTERNOON

Exhibit Hall A2

12:30 Voting on Awards for Best Presentation and Poster and Presentation of Awards

1:00 MATHEMATICS, COMPUTER, AND INFORMATION SCIENCES ALUMNI SURVEY ONLINE FORM

John Stepney, Mississippi Valley State University, Itta Bena, MS 38941

In this project an online MCIS alumni survey form is created that connects to an Oracle database and automatically creates an online report. The MCIS alumni survey allows the alumni of Mississippi Valley State University's Mathematics, Computer, and Information Sciences department alumni to complete the survey and submit their data online. The database is created using Oracle database management system and the Structure Query Language. After the information enters the Oracle database then an online report is automatically generated based on the data from the database. The purpose of this project is to allow the MCIS department to generate summative data about the department and its graduates. The MCIS online survey will make retrieving the alumni's information faster and more efficient. The MCIS alumni survey will be transferred to the Mississippi Valley State University server to allow MCIS alumni to access the form off campus.

Printer friendly Cite/link Email Feedback | |

Publication: | Journal of the Mississippi Academy of Sciences |
---|---|

Date: | Jan 1, 2006 |

Words: | 2889 |

Previous Article: | Marine and atmospheric sciences. |

Next Article: | Physics and engineering. |