Digital Commons Network^™

A Multithreaded Message-Passing System For High Performance Distributed Computing Applications, Sung-Yong Park, Joohan Lee, Salim Hariri

Electrical Engineering and Computer Science - All Scholarship

High Performance Distributed Computing (HPDC) applications require low-latency and high-throughput communication services and HPDC applications have different Quality of Service (QOS) requirements (e.g., bandwidth requirement, flow/error control algorithms, etc.). The communication services provided by traditional message-passing systems are fixed and thus can not be changed to meet the requirements of different HPDC applications. NYNET (ATM wide area network testbed in New York state) Communication System (NCS) is a multithreaded message-passing system developed at Syracuse University that provides high-performance and flexible communication services. In this paper, we overview the general architecture of NCS and present how NCS communication services are implemented. …

A Problem Solving Environment For Network Computing, Salim Hariri, Haluk Topcuoglu, Wojtek Furmanski, Dongmin Kim, Yoonhee Kim

Electrical Engineering and Computer Science - All Scholarship

The current advances in high-speed networks and WWW technologies have made network computing a cost-effective high performance computing environment. New software development models and problem solving environments must be developed to utilize the network computing environment efficiently. In this paper we present Virtual Distributed Computing Environment (VDCE), which provides a problem solving environment for high-performance distributed computing over wide-area networks. VDCE enables scientists to develop distributed applications without knowing the detailed architecture of the underlying resources. VDCE provides well-defined library functions that relieve end users from tedious task implementations and it supports software reusability. The VDCE software architecture consists of …

Automatic Granularity Control For Load Balancing Of Concurrent Particle Simulations, Marc Rieffel, Stephen Taylor, Jerrell Watts

Electrical Engineering and Computer Science - All Scholarship

This paper demonstrates the use of automatic granularity control as part of dynamic load balancing for irregular, particle-based simulations. Performance optimization techniques are considered in the context of a concurrent Direct Simulation Monte Carlo method used to study the rarefied gas flow inside three-dimensional plasma reactors. Several computational techniques are used to reduce the overall time to deliver realistic threedimensional results. The effectiveness of dynamic load balancing and granularity control are presented for large-scale simulations on distributed-memory multicomputers.

Performance Enhancement Using Intra-Server Caching In A Continuous Media Server, Chutimet Srinilta, Alok Choudhary

Electrical Engineering and Computer Science - All Scholarship

Continuity of stream playback is the crucial constraint in designing a continuous media server. From a distributed memory architectural model developed earlier, we found that there were many points where the stream capacity of the server could be improved. The stream capacity was usually limited by the storage bottlenecks. Serving streams from memory cache eliminates disk accesses and data transfers between nodes which, in turn, helps relieve those bottlenecks. However, the capacity of the server ultimately depends on client access pattern. Client request assignment has an impact on cache hit ratio as well as workload distribution. It is also the …

Verification And Validation Of Simulation Models, Robert G. Sargent

Electrical Engineering and Computer Science - All Scholarship

This paper discusses verification and validation of simulation models. The different approaches to deciding model validity are presented; how model verification and validation relate to the model development process are discussed; various validation techniques are defined; conceptual model validity, model verification, operational validity, and data validity are described; ways to document results are given; and a recommended procedure is presented.

Skew-Insensitive Parallel Algorithms For Relational Join, Khaled Alsabti, Sanjay Ranka

College of Engineering and Computer Science - Former Departments, Centers, Institutes and Projects

Join is the most important and expensive operation in relational databases. The parallel join operation is very sensitive to the presence of the data skew. In this paper, we present two new parallel join algorithms for coarse grained machines which work optimally in presence of arbitrary amount of data skew. The first algorithm is sort-based and the second is hash-based. Both of these algorithms employ a preprocessing phase (prior to the redistribution phase) to equally partition the work among the processors. The proposed algorithms have been designed for memory resident-data. However, they can be extended to disk resident-data. These algorithms …

The Design And Evaluation Of A Virtual Distributed Computing Environment, Haluk Topcuoglu, Salim Hariri, Dongmin Kim, Yoonhee Kim, Xue Bing

Electrical Engineering and Computer Science - All Scholarship

In this paper we present the Virtual Distributed Computing Environment (VDCE), a metacomputing environment currently being developed at Syracuse University. VDCE provides an efficient web-based approach for developing, evaluating and visualizing large-scale distributed applications that are based on predefined task libraries on diverse platforms. The VDCE task libraries relieve end-users of tedious task implementations and also support reusability. The VDCE software architecture is described in terms of three modules: a) the Application Editor, a user-friendly application development environment that generates the Application Flow Graph (AFG) of an application; b) the Application Scheduler, which provides an efficient task-to-resource mapping of AFG; …

Adaptive Linkage Crossover, Ayed A. Salman, Kishan Mehrotra, Chilukuri K. Mohan

Electrical Engineering and Computer Science - All Scholarship

Problem-specific knowledge is often implemented in search algorithms using heuristics to determine which search paths are to be explored at any given instant. As in other search methods, utilizing this knowledge will more quickly lead a genetic algorithm (GA) towards better results. In many problems, crucial knowledge is not found in individual components, but in the interrelations between those components. For such problems, we develop an interrelation (linkage) based crossover operator that has the advantage of liberating GAs from the constraints imposed by the fixed representations generally chosen for problems. The strength of linkages between components of a chromosomal structure …

Jworb - Java Web Object Request Broker For Commodity Software Based Visual Data Ow Metacomputing Programming Environment, Geoffrey C. Fox, Wojtek Furmanski, Hasan T. Ozdemir

Northeast Parallel Architecture Center

Programming environments and tools that are simultaneously sustainable, highly functional, robust and easy to use have been hard to come by in the HPDC area. This is partially due to the difficulty in developing sophisticated customized systems for what is relatively small part of the worldwide computing enterprise. As the commodity software becomes naturally distributed with the onset of Web and Intranets, we observe now a new trend in HPDC community [1, 8, 12] to base high performance computing on the modern enterprise computing technologies. .. JWORB is a multi-protocol Java server under development at NPAC, currently capable of handling …

Language Bindings For A Data-Parallel Runtime, Bryan Carpenter, Geoffrey C. Fox, Donald Leskiw, Xinying Li, Yuhong Wen

Northeast Parallel Architecture Center

The NPAC kernel runtime, developed in the PCRC (Parallel Compiler Runtime Consortium) project, is a runtime library with special support for the High Performance Fortran data model. It provides array descriptors for a generalized class of HPF-like distributed arrays, support for parallel access to their elements, and a rich library of collective communication and arithmetic operations for manipulating these arrays. The library has been successfully used as a component in experimental HPF translation systems. With prospects for early appearance of fully-featured, efficient HPF compilers looking questionable, we discuss a class of more easily implementable data-parallel language extensions that preserve many …

Java/Corba Based Real-Time Infrastructure To Integrate Event-Driven Simulations, Collaboration And Distributed Object/Componentware Computing, Geoffrey C. Fox, Wojtek Furmanski, Hasan T. Ozdemir

Northeast Parallel Architecture Center

We are discussing the four major standard candidates for distributed object/componentware computing: Java, CORBA, COM and WOM within our proposed coordination framework we call Pragmatic Object Web (POW). We describe our integration approach based on multi-protocol middleware server JWORB (Java Web Object Request Broker) that currently integrates HTTP and IIOP and which we now further develop to also support COM and WOM core functionalities. We are also experimenting with visual data ow authoring front-ends using NPAC WebFlow system on top of JWORB based software bus. Finally, we illustrate our technologies in one major application domain- DoD Modeling and Simulation- where …

Towards A Java Environment For Spmd Programming, Bryan Carpenter, Guansong Zhang, Geoffrey C. Fox, Xiaoming Li

Northeast Parallel Architecture Center

As a relatively straightforward object-oriented language, Java is a plausible basis for a scientific parallel programming language. We outline a conservative set of language extensions to support this kind of programming. The programming style advocated is Single Program Multiple Data (SPMD), with parallel arrays added as language primitives. Communications involving distributed arrays are handled through a standard library of collective operations. Because the underlying programming model is SPMD programming, direct calls to other communication packages are also possible from this language.

Considerations In Hpjava Language Design And Implementation, Guansong Zhang, Bryan Carpenter, Geoffrey C. Fox, Xinying Li, Yuhong Wen

Northeast Parallel Architecture Center

This paper discusses some design and implementation issues in the HPJava language. The language is briefly reviewed, then the class library that forms the foundation of the translation scheme is described. Through example codes, we illustrate how HPJava source codes can be translated straightforwardly to ordinary SPMD Java programs calling this library. This is followed by a discussion of the rationale for introducing the language in the first place, and of how various language features have been designed to facilitate efficient implementation.

A Global Communication Optimization Technique Based On Data-Flow Analysis And Linear Algebra, Mahmut Kandemir, P. Banerjee, Alok Choudhary, J. Ramanujam

Electrical Engineering and Computer Science - All Scholarship

Reducing communication overhead is extremely important in distributed-memory message-passing architectures. In this paper, we present a technique to improve communication that considers data access patterns of the entire program. Our approach is based on a combination of traditional data-flow analysis and a linear algebra framework, and works on structured programs with conditional statements and nested loops but without arbitrary goto statements. The distinctive features of the solution are the accuracy in keeping communication set information, support for general alignments and distributions including block-cyclic distributions and the ability to simulate some of the previous approaches with suitable modifications. We also show …

Clouds: A Decision Tree Classifier For Large Datasets, Khaled Alsabti, Sanjay Ranka, Vineet Singh

Electrical Engineering and Computer Science - All Scholarship

Classification for very large datasets has many practical applications in data mining. Techniques such as discretization and dataset sampling can be used to scale up decision tree classifiers to large datasets. Unfortunately, both of these techniques can cause a significant loss in accuracy. We present a novel decision tree classifier called CLOUDS, which samples the splitting points for numeric attributes followed by an estimation step to narrow the search space of the best split. CLOUDS reduces computation and I/O complexity substantially compared to state of the art classifiers, while maintaining the quality of the generated trees in terms of accuracy …

Steady State Memetic Algorithm For Partial Shape Matching, Ender Ozcan, Chilukuri K. Mohan

Electrical Engineering and Computer Science - All Scholarship

Shape matching techniques are important in machine intelligence, especially in applications such as robotics. Currently, there are three major approaches to shape recognition: statistical, syntactic and neural approaches. This paper presents a fourth approach: evolutionary algorithms. A steady state memetic algorithm is shown to be successful in matching shapes even when they are partially obscured, and even in the presence of noise in the input image.

Integrating Security Into The Curriculum, Cynthia E. Irvine, Shiu-Kai Chin, Deborah Frincke

Electrical Engineering and Computer Science - All Scholarship

The number of skilled practitioners of computer security who are able to address the complexities of modern technology and are familiar with successful approaches to system security is very small. People want security but are faced with two difficulties. First, they do not know how to achieve it in the context of their enterprises. They may not even know of a way to translate organizational procedures into policies, much less implement a set of mechanisms to enforce those policies. Second, they have no way of knowing whether their chosen mechanisms are effective. The recent US Presidential Commission on Critical Infrastructure …

Secure Delegation For Distributed Object Environments, Nataraj Nagaratnam, Doug Lea

Electrical Engineering and Computer Science - All Scholarship

SDM is a Secure Delegation Model for Java-based distributed object environments. SDM extends current Java security features to support secure remote method invocations that may involve chains of delegated calls across distributed objects. The framework supports a control API for application developers to specify mechanisms and security policies surrounding simple or cascaded delegation. Delegation may also be disabled and optionally revoked. These policies may be controlled explicitly in application code, or implicitly via administrative tools.

Search For Evidence Of Compositeness At Lep I, R. Barate, M. Thulasidas

Research Collection School Of Computing and Information Systems

Extensive searches for evidence of compositeness of quarks and leptons have been made in the complete set of ALEPH data from LEP running at the Z peak. These include searches for radiative and weak decays of excited fermions, produced either in pairs or singly, and a scalar partner of the Z. Coupling and branching ratio limits deduced from these searches are presented.

Automatic Granularity Control For Load Balancing Of Concurrent Particle Simulations, Marc Rieffel, Stephen Taylor, Jerrell Watts

Electrical Engineering and Computer Science - All Scholarship

This paper demonstrates the use of automatic granularity control as part of dynamic load balancing for irregular, particle-based simulations. Performance optimization techniques are considered in the context of a concurrent Direct Simulation Monte Carlo method used to study the rarefied gas flow inside three-dimensional plasma reactors. Several computational techniques are used to reduce the overall time to deliver realistic threedimensional results. The effectiveness of dynamic load balancing and granularity control are presented for large-scale simulations on distributed-memory multicomputers.

An Upper Limit On The Τ Neutrino Mass From Three- And Five-Prong Tau Decays, R. Barate, M. Thulasidas

Research Collection School Of Computing and Information Systems

A bound on the tau neutrino mass is established using the data collected from 1991 to 1995 at s√≃mZ with the ALEPH detector. Two separate limits are derived by fitting the distribution of visible energy vs invariant mass in τ−→2π−π+ντ and τ−→3π−2π+(π0)ντ decays. The two results are combined to obtain a 95% confidence level upper limit of 18.2 MeV/c2 on the mass of the tau neutrino.

Observation Of Doubly-Charmed B Decays At Lep, R. Barate, M. Thulasidas

Research Collection School Of Computing and Information Systems

A search for doubly-charmed B decays with both charmed mesons reconstructed is performed, using about 3.8 million hadronic Z decays recorded with the ALEPH detector at LEP. A clear signal is observed in the channels B→DsD¯(X)B→DsD¯(X) and B→DD¯(X)B→DD¯(X) (where D can be either a D00, a D++ or a D∗+∗+), providing the first direct evidence for doubly-charmed B decays involving no DsDs production. Evidence for associated K0SKS0 and K±K± production in the decays B→DD¯(X)B→DD¯(X) is also presented and some candidates for completely reconstructed decays B→DsD¯(nπ)B→DsD¯(nπ), B→DD¯K0SB→DD¯KS0 and B→DD¯K±B→DD¯K± are observed. Furthermore, candidates for the two-body Cabibbo suppressed decays B0→D∗−D∗+B0→D∗−D∗+ and …

Searches For Charginos And Neutralinos In E+E- Collisions At √S = 161 And 172 Gev, R. Barate, M. Thulasidas

Research Collection School Of Computing and Information Systems

The data recorded by the ALEPH detector at centre-of-mass energies of 161, 170, and 172 GeV are analysed for signals of chargino and neutralino production. No evidence of a signal is found, although candidate events consistent with the expectations from Standard Model processes are observed. Limits at 95% C.L. on the production cross sections are derived and bounds on the parameters of the Minimal Supersymmetric Standard Model are set. The lower limit on the mass of the lightest ch is 85.5 GeV/c2GeV/c2 for gaugino-like charginos (μ=−500GeV/c2μ=−500GeV/c2), and 85.0 GeV/c2GeV/c2 for Higgsino-like charginos (M2=500GeV/c2M2=500GeV/c2), for heavy sneutrinos (Mν~≥200GeV/c2Mν~≥200GeV/c2) and tanβ=2–√tan⁡β=2. The …

Aegis Data Analysis And Reduction (Adar) In Support Of The Aegis Weapon System (Aws), June Bullard Gaines

Theses and Dissertations

The AEGIS Weapons System (AWS), part of the AEGIS Combat System (ACS), is an integral part of the defense system on U.S. Navy AEGIS-class ships. AEGIS Data Analysis and Reduction (ADAR) has been developed to assist in the evaluation of the AWS data. ADAR, along with the AWS and ACS, has evolved through the years to accommodate advances in technology and computer programming languages. Additionally, ADAR has evolved so that users located at sites other than the Naval Surface Warfare System Dahlgren Division (NSWCDD), Dahlgren, Virginia, can reduce tactical system data and perform data analysis using the reduced data.

This …

Studies Of Quantum Chromodynamics With The Aleph Detector, R. Barate, M. Thulasidas

Research Collection School Of Computing and Information Systems

No abstract provided.

Evaluation Of Information Modeling Methods: A Review, Keng Siau, Matti Rossi

Research Collection School Of Computing and Information Systems

Hundreds of modeling methods are in existence today. Yet practitioners and researchers are zealously "producing" new modeling methods. The "blooming" of modeling methods is not the problem; the lack of standardize techniques for evaluating them is. To further complicate the matter, most of the modeling methods are introduced based on common sense and intuition. Theoretical foundations and empirical evidence are severely lacking. With the current state of affairs, evaluation of modeling methods has become necessary. Comparing modeling methods provides us with the necessary knowledge and understanding on the strengths and weaknesses of each method. This knowledge can also guide us …

An Efficient Library For Parallel Ray Tracing And Animation, John Edward Stone

Masters Theses

"A parallel ray tracing library is presented for rendering high detail images of three dimensional geometry and computational fields. The library has been developed for use on distributed memory and shared memory parallel computers and can also run on sequential computers. Parallelism is achieved through the use of message passing and threads. It is shown that the library achieves almost linear scalability when run on large distributed memory parallel computers as well as large shared memory parallel computers.

Several applications of parallel rendering are explored including rendering of CAD models, animation, magnetic resonance imaging, and visualization of volumetric flow fields. …

Measurement Of The Fraction Of Hadronic Z Decays Into Charm Quark Pairs, R. Barate, M. Thulasidas

Research Collection School Of Computing and Information Systems

The full statistics of hadronic Z decays collected with the ALEPH detector are analysed to measure, by three methods, the ratio, RcRc, of the partial decay width of the Z into cc¯cc¯ quarks to the total hadronic width. One method uses the inclusive pp and p⊥p⊥ spectra of electrons. The other two use a double-tagging technique to measure RcRc independently of the charm-tagging efficiency. In one, the tagging procedure uses the reconstruction of high momentum charmed mesons (D∗+D∗+, D+D+ and \({\rm D^0}$) to detect the charm quarks. In the other, a charm quark is tagged from the D∗D∗'s reconstructed in …

Evaluating Information Modeling Methods – A Review, Keng Siau, M. Rossi

Research Collection School Of Computing and Information Systems

Hundreds of modeling methods are in existence today, yet practitioners and researchers are zealously “producing” new modeling methods. The “blooming” of modeling methods is not the problem; the lack of standardized techniques for evaluating them is. To further complicate the matter, most of the modeling methods are introduced based on common sense and intuition. Theoretical foundations and empirical evidence are severely lacking. With the current state of affairs, evaluation of modeling methods has become necessary. Comparing modeling methods provides us with the necessary knowledge and understanding on the strengths and weaknesses of each method. This knowledge can also guide us …

A Transaction Model For Mobile Computing, Sanjay Kumar Madria, Bharat Bhargava

Computer Science Faculty Research & Creative Works

We introduce a prewrite operation before a write operation in a mobile transaction to improve data availability. A prewrite operation does not update the state of a data object but only makes visible the value that the data object will have after the commit of the transaction. Once the transaction has read all the values and declares all the prewrites, it can precommit at a mobile host. The remaining transaction's execution is shifted to the stationary host. Writes on a database consume both time and resources at the stationary host and are therefore delayed. A pre-committed transaction's prewrite values are …