1998  Graduate Student Research Conference Program (GSRC98)


KEYNOTE SPEAK  - John Carolan

Long-Lived Design

This talk looks at factors that contribute to a longer life for software designs.  How well do people understand a design?  Is their understanding restricted by current fashions?  Or is it more universal, appealing to a wide audience over the long term?  Is it natural and brief to learn?  How well is it promoted?



 

Controlled Simplification of Genus for Polygonal Models

Jihad El-Sana

Genus-reducing simplifications are important in constructing multiresolution hierarchies for level-of-detail-based rendering,  especially for datasets that have several relatively small holes, tunnels, and cavities. We present a genus-reducing simplification approach that is complementary to the existing work on genus-preserving simplifications. We propose a simplification framework in which genus-reducing and genus-preserving simplifications alternate to yield much better multiresolution hierarchies than would have been possible by using either one of them.  In our approach we first identify the holes and the concavities by extending the concept of alpha-hulls to polygonal meshes under the L-infinty distance metric and then generate valid triangulations to fill them.


The Cube-4 Real-Time Volume Rendering Architecture

Ingmar Bitter

We describe our high-performance special purpose hardware system, Cube-4, for displaying and manipulating high resolution volumetric data sets in real-time. Cube-4 is designed to render 512x512x512, 16bit per voxel, data sets at 30 frames per second. Cube-4 implements a ray-casting algorithm in a highly-parallel and pipelined architecture. We also describe the recent Cube4 derivative Cube-4Light, which simplified the design but also added hardware support for perspective projections.


Implementation and Evaluation of Parallel Mesa Library

Tulika Mitra

  In this talk, we describe the implementation and performance evaluation
  of a 3D graphics library that can be readily linked with
  parallel applications to provide run-time visualization
  on large-scale message-passing parallel machines, such as
  Intel Paragon. The prototype implementation is currently fully operational,
  and is based on a public-domain OpenGL implementation Mesa,
  and a sort-last  parallelization strategy.
  Through a detailed performance analysis, we show that the scalability
  of the current prototype is close to the theoretical limit for
  the given hardware architecture, and that further performance improvement
  on the same platform is unlikely to change the scalability characteristics,
  especially when the number of rendering processors involved is large.
  We also develop a unified framework to describe
  parallel compositing algorithms and show that two popular parallel compositing
  algorithms, binary swapping and parallel pipeline compositing,
  are just two extreme instances of this framework. Such a framework will
  allow users to tailor the compositing algorithm according to the
  computation/communication characteristics of specific parallel machines by tuning the
  parameters appropriately. The current parallel Mesa library prototype implements
  such a parameterizable family of compositing algorithms.


Physically-based Animation of Volumetric Objects

Yan Chen

 
This paper presents a voxel-based animation technique which employs
  either a mass-spring model or a finite element model.  Two volumetric
  objects, a voxelized chair and a scanned muscle, are used as case
  studies with the different models.  The mass-spring model is used to
  show an animation sequence of a falling and bouncing chair.  Wireframe
  display and volume rendering are used to display a real-time animation
  of the process.  In addition, a real-time simulation is carried out by
  the Finite Element Method (FEM) of a voxel-based multi-resolution
  muscle mesh.  Two techniques, a direct integration and a simplified
  modal analysis method are discussed in the context of applying FEM for
  muscle deformation.  Local deformation optimization with modal
  analysis for higher resolution muscle volumetric animation, which
  allows accurate prediction of muscle deformation changes, has been
  used.  Physiological muscle force has been considered and a
  biomechanically-based 3D FEM muscle model has been implemented.
  Realistic animations have been produced based on the FEM simulation
  with various graphics techniques.



 

Real Time Performance Guarantees over Wired/Wireless LANs
 
Prashant Pradhan

Distributed audio and video applications require end-to-end
  performance guarantees from the underlying networks for smooth
  delivery of real-time media streams.  The end-to-end nature dictates
  that real-time support be built into both wide-area and local-area
  networks. With the enormous growth of portable computers equipped with
  networking capabilities, it is important to develop real-time
  protocols for wireless LAN technologies, such that real-time
  audio/video communication services can be extended to these mobile
  terminals as well. This paper describes the implementation effort of
  supporting end-to-end performance guarantees over a heterogeneous
  local-area network comprising of both wired and wireless LAN segments.
  This work also addresses the performance guarantee issues associated
  with terminal mobility in the wireless segments.


Development of Analytical and Simulation models for a Reliable Multicast Transport Protocol

Gursel Taskale
 


Model checking the i-protocol

Yifei Dong

 Model checking is a verification technique aimed at determining whether a system specification possesses a property expressed as
 a temporal logic formula.  The i-protocol is a bidirectional sliding-window protocol employed in the GNU uucp file transfer utility.  I'll report how we verified the i-protocol and found a non-trivial livelock error in it, as an example to introduce the conception, success, and challenge of model checking.


Correct Execution at Different isolation Level

Shiyong Lu

  The ANSI-92 specifications defines four levels of isolation:
  READ UNCOMMITTED, READ COMMITTED, REPEATABLE READ and SERIALIZABLE.
  One additional important isolation level, SNAPSHOT isolation, is
  also very popular in nowaday's commercial dababase managers.
  In general, the weaker the isolation level, the greater the degree of
  concurrency allowed. However, only the SERIALIZABLE level provides correct
  behavior for all possible applications.

  We believe that, for performance reasons, designers will continue to execute
  applications at isolation levels lower than serializable. Our study of
  such non-serializable schedules addresses two issues:
   1. We would like to develop a technique for determining, for a given
      application, the least restrictive isolation level consistent with
      its semantics correctness. We are also interested in the more
      challenging problem of allowing different transactions in the
      same application to run at different isolation levels.

   2. In situations in which the semantics correctness of an application
      is not supported at a particular isolation level, we want to
      investigate techniques that can be used within the transactions
      themselves to provide additional synchronization necessary to achieve
      semantics correctness at that level.

  In this talk, we present part of our research results for the first issue.
  Specifically, we give semantics conditions for different isolation levels
  such that, if a set of transactions satisfy the semantics condition for
  a particular isolation level, then we can run the set of transactions at
  that isolation level while semantics correctness is guaranteed. Formal
  proofs for these theorems are available upon request.


WebFlow Logic - A Logic for Everyday Workflow Practitioner
 
Hasan Davulcu

A workflow can simply be defined as a set of tasks that cooperate to implement
  a business process. Executing a workflow thus involves coordinated execution
  of multiple long-running activities in an environment of distributed,
  heterogeneous processing entities.

  We propose Concurrent Transaction Logic (CTR) as a declarative language for
  specifying, analyzing, and scheduling of workflows.

  At present, we have a prototypical web-enabled system, WebFlow Logic (WFL),
  implemented in XSB (an efficient deductive database system).

  The WFL system is tolerant to crashes and the workflow instances in progress
  are forward recoverable. CA's object-oriented database, Jasmine is used as
  the workflow database for storing and monitoring the progress of active
  workflow instances.



 

Tabled Constraint Logic Programming over Finite Domains
 
Baoqiu Cui

Constraint Logic Programming (CLP) is a new class of programming
  languages which extend the declarativity of logic programming by the
  addition of constraint solving.  In the past decade, CLP has been
  studied deeply in different directions and on several computation
  domains.  Just like in Logic Programming, tabling plays an important
  role in CLP.  A CLP system with tabling techniques can not only avoid
  redundant computations and many infinite loops, but also give an easy
  way to find the optimal solutions for some problems.  This talk will
  discuss some issues in applying tabling techniques into CLP over finite
  domains.  An meta-interpreter of a CLP language over finite domains,
  which is written based on the XSB system, will be introduced.
 



 
Compressed domain text retrieval system on the internet

Lan Huang

 The topic is a compressed domain text retrieval system on the internet, which supports approximate string search,efficient DB update, and also allows time/space tradeoff .This system can be extended to other applications like video annotation retrieval,technical report search.



 

A Multimedia Communication System
 
Sergio Silva

The basic idea of the Multimedia Communication System is to extend the
  functionality of Internet Relay Chat (IRC) systems to include other types
  of media, such as voice and sketch image, to the usual text interface.
  The goal is to create a virtual study room, to allow students, TAs and
  professors to interactively discuss class related material through
  communications networks, without being physically present.


Zodiac: A history-based video authoring system
 
Chuan-Kai Yang

Easy-to-use audio/video authoring tools play a crucial role in moving
  multimedia programs from research curiosity to main-stream applications.
  Recently we implement an interactive
  video authoring system called Zodiac, which employs an innovative
  edit history abstraction to support several unique editing features
  not found in existing commercial and research video editing systems.
  Zodiac provides users a conceptually clean and semantically
  powerful branching history model of edit operations to organize the
  authoring process, and to navigate among the design alternatives.
  In addition, by analyzing the edit history, Zodiac is able
  to reliably detect a composed stream's shot and scene boundaries,
  which facilitate interactive video browsing. Zodiac also features
  a video object annotation capability that allows users to
  associate annotations to moving objects in a video sequence.
  The annotations themselves could be text, image, audio, or video.
  In this talk I will briefly talk about these three features in our
  system: branching history, shot/scene detection, video object annotation.



 

EtheReal: A Framework for Bandwidth Guarantees over Switched Ethernet Network

Srinidhi Varadarajan

  With advancement of technology and availability of powerful desktop
  computers, human computer interaction using more natural data sources like
  video and audio has become commonplace. These multimedia sources require
  bandwidth guarantees from the network on which they travel. Specifically,
  multimedia data need to get from the source to the destination in real time,
  or within a predictable period of time. This is in contrast to conventional
  computer network traffic, such as file transfers or email, which have no
  stringent timing constraints. However, the most common LAN technology in use
  today, viz. Ethernet, treats both realtime and non-real-time data
  sources equally, and thus does not match the networking requirements of
  distributed multimedia applications.

  The goal of the EtheReal project is to design and implement a scalable Fast
  Ethernet Switch that provides real time bandwidth guarantees in a way that
  is transparent to the host machines on the network.  The switch isolates
  real-time sources from both non-real-time sources and other real time
  sources. It also prioritizes real time sources based upon their requirements
  from the network.  A distinguishing feature of the work is the capability to
  provide these guarantees without any change to the existing operating
  system software or hardware on the hosts connected to the network. The main
  contribution of this research is the development of a novel real-time
  connection setup mechanism that assumes only common TCP/IP protocol support
  and is thus portable across multiple OS platforms.

  An operational EtheReal prototype has been built using an off-the-shelf PC
  and commodity Ethernet cards. The hosts on the network can run different
  operating systems such as Windows 95, Linux and FreeBSD.  We are currently
  working on fault tolerance mechanisms to continue switch operation across
  network hardware failures. The other major area of future research is in
  supporting real-time multicast, which is the enabling technology for
  Internet push technologies.


Performance Evaluation of An RSVP-Capable Router

Anindya Neogi

  RSVP is a bandwidth reservation protocol that allows distributed real-time
  applications such as video-conferencing software to make bandwidth
  reservations over packet-switched networks. Coupled with real-time
  scheduling mechanisms built into packet routers, the network guarantees
  to provide the reserved bandwidth throughout the lifetime of the
  applications. Although guaranteed services are of great value
  to both end users and carrier providers, their performance cost, due
  to additional control and data processing overhead, can potentially
  have a negative impact on the packet throughput and latency
  of the RSVP-capable routers. The goal of this paper is to examine
  the performance cost of RSVP based on measurements from an
  industry-strength RSVP implementation on a commercial IP router.
  The focus is on the detailed evaluation of the performance implications
  of various architectural decisions in RSVP.
  We found that RSVP's control messages do not incur significant
  overhead in terms of processing delay and bandwidth consumption.
  However, the performance overhead of real-time packet scheduling is
  noticeable in the presence of a large number of real-time
  connections. In extreme cases, the performance guarantees of
  existing real-time connections may not be kept, and some best-effort
  packets are actually dropped, although the overall bandwidth
  requirement from these connections is smaller than the available
  link bandwidth.
 
 
 

Return to the GRC home page


If you have problems with this page, please send E-mail to:
evans@cs.sunysb.edu