Joe Hoffert, Douglas Schmidt
Citation
Joe Hoffert, Douglas Schmidt. "FLEXible Middleware And Transports (FLEXMAT) for Real-time Event Stream Processing (RT-ESP) Applications". Talk or presentation, July, 2009.
Abstract
FLEXible Middleware And Transports (FLEXMAT) for Real-time Event
Stream Processing (RT-ESP) Applications
Real-time Event Stream Processing (RT-ESP) applications support mission
critical systems (such as collaboration of weather monitoring radars to
predict life-threatening weather) by managing and coordinating multiple
streams of event data that have (possibly distinct) timeliness
requirements. Streams of event data may originate from sensors (e.g.,
surveillance cameras, temperature probes), as well as other types of
monitors (e.g., online stock trade feeds).
Real-time event notification middleware provides a promising platform
for RT-ESP applications. This middleware requires scalability and
flexibility to function effectively in large-scale environments. No
single transport protocol provides sufficient scalability, flexibility,
and real-time QoS properties. R&D is therefore needed to determine how
existing and new transport protocols can be configured and tuned to meet
the requirements of real-time event notification middleware in
large-scale resource-constrained environments.
For example, a search-and-rescue application used to detect and locate
survivors as part of disaster relief efforts needs to coordinate and
synchronize a video data stream from one platform (e.g., a videocamera
mounted atop a building) and a thermal scan data stream from another
platform (e.g., a thermal imaging camera attached to an unmanned aerial
vehicle). These data streams are then fused together to detect
survivors. The SAR application must do the processing of multiple event
data streams while operating in less than ideal conditions due to the
nature of the disaster.
Moreover, there are not only multiple senders of event stream data, but
there are also potentially many receivers of the data, e.g., a
helicopter performing rescue operations, local disaster relief
headquarters, and Federal Emergency Management Agency (FEMA)
headquarters. Other applications with similar data stream requirements
include tracking the prices of multiple stocks, synchronizing medical
telemetry data in a wireless medical emergency room, and monitoring
global weather via data streams from multiple sensors.
Various transport protocols can affect the performance of real-time
event notification systems particularly for RT-ESP applications in
resource constrained or turbulent environments. Systematic evaluations
are generally lacking, however, of transport protocols integrated with
real-time event notification middleware. These evaluations help provide
insight into the benefits and drawbacks of various transport protocols
for a particular configuration or environment. These evaluations are
thus valuable not only for standard protocols (such as TCP, UDP, and IP
multicast) but also for custom protocols from both industry and
academia. Custom protocols can provide advantages over standard
protocols for certain QoS requirements or environment configurations.
Empirical measurements and evaluations of both standard and custom
protocols integrated with real-time event notification systems provide
guidance as to the selection of a transport protocol for a given
environment and QoS requirements.
We have integrated the Adaptive Network Transport (ANT) framework with
two implementations of the Data Distribution Service (DDS). The
integration with the DDS implementations is not meant for performance
comparisons between DDS products, but to showcase different integration
approaches and also to generalize experimental results of the impact of
transport protocols across DDS implementations. The ANT framework we
support is based on extensions to the Ricochet transport protocol.
Ricochet is a scalable reliable multicast protocol that combines high
data rates with strong probabilistic delivery guarantees. Ricochet
balances the need for reliability and low latency while providing
fine-grained tunability with regard to reliability, latency, and network
bandwidth usage.
We will characterize the performance of various transport protocols,
including Ricochet, the standard DDS Interoperability (DDSI) protocol,
TCP, UDP, and IP multicast. We also will present experimental results
that quantify the trade-offs of the various transport protocols in a
range of representative environments using DDS.
Electronic downloads
- ADAMANT-RTWS09.pptx · application/vnd.openxmlformats-officedocument.presentationml.pre · 5119 kbytes
| Citation formats |
- HTML
Joe Hoffert, Douglas Schmidt. <a href="http://www.truststc.org/pubs/675.html" ><i>FLEXible Middleware And Transports (FLEXMAT) for Real-time Event Stream Processing (RT-ESP) Applications</i></a>, Talk or presentation, July, 2009.
- Plain text
Joe Hoffert, Douglas Schmidt. "FLEXible Middleware And Transports (FLEXMAT) for Real-time Event Stream Processing (RT-ESP) Applications". Talk or presentation, July, 2009.
- BibTeX
@presentation{HoffertSchmidt09_FLEXibleMiddlewareTransportsFLEXMATForRealtimeEvent, author = {Joe Hoffert and Douglas Schmidt}, title = {FLEXible Middleware And Transports (FLEXMAT) for Real-time Event Stream Processing (RT-ESP) Applications}, month = {July}, year = {2009}, abstract = {FLEXible Middleware And Transports (FLEXMAT) for Real-time Event Stream Processing (RT-ESP) Applications Real-time Event Stream Processing (RT-ESP) applications support mission critical systems (such as collaboration of weather monitoring radars to predict life-threatening weather) by managing and coordinating multiple streams of event data that have (possibly distinct) timeliness requirements. Streams of event data may originate from sensors (e.g., surveillance cameras, temperature probes), as well as other types of monitors (e.g., online stock trade feeds). Real-time event notification middleware provides a promising platform for RT-ESP applications. This middleware requires scalability and flexibility to function effectively in large-scale environments. No single transport protocol provides sufficient scalability, flexibility, and real-time QoS properties. R\&D is therefore needed to determine how existing and new transport protocols can be configured and tuned to meet the requirements of real-time event notification middleware in large-scale resource-constrained environments. For example, a search-and-rescue application used to detect and locate survivors as part of disaster relief efforts needs to coordinate and synchronize a video data stream from one platform (e.g., a videocamera mounted atop a building) and a thermal scan data stream from another platform (e.g., a thermal imaging camera attached to an unmanned aerial vehicle). These data streams are then fused together to detect survivors. The SAR application must do the processing of multiple event data streams while operating in less than ideal conditions due to the nature of the disaster. Moreover, there are not only multiple senders of event stream data, but there are also potentially many receivers of the data, e.g., a helicopter performing rescue operations, local disaster relief headquarters, and Federal Emergency Management Agency (FEMA) headquarters. Other applications with similar data stream requirements include tracking the prices of multiple stocks, synchronizing medical telemetry data in a wireless medical emergency room, and monitoring global weather via data streams from multiple sensors. Various transport protocols can affect the performance of real-time event notification systems particularly for RT-ESP applications in resource constrained or turbulent environments. Systematic evaluations are generally lacking, however, of transport protocols integrated with real-time event notification middleware. These evaluations help provide insight into the benefits and drawbacks of various transport protocols for a particular configuration or environment. These evaluations are thus valuable not only for standard protocols (such as TCP, UDP, and IP multicast) but also for custom protocols from both industry and academia. Custom protocols can provide advantages over standard protocols for certain QoS requirements or environment configurations. Empirical measurements and evaluations of both standard and custom protocols integrated with real-time event notification systems provide guidance as to the selection of a transport protocol for a given environment and QoS requirements. We have integrated the Adaptive Network Transport (ANT) framework with two implementations of the Data Distribution Service (DDS). The integration with the DDS implementations is not meant for performance comparisons between DDS products, but to showcase different integration approaches and also to generalize experimental results of the impact of transport protocols across DDS implementations. The ANT framework we support is based on extensions to the Ricochet transport protocol. Ricochet is a scalable reliable multicast protocol that combines high data rates with strong probabilistic delivery guarantees. Ricochet balances the need for reliability and low latency while providing fine-grained tunability with regard to reliability, latency, and network bandwidth usage. We will characterize the performance of various transport protocols, including Ricochet, the standard DDS Interoperability (DDSI) protocol, TCP, UDP, and IP multicast. We also will present experimental results that quantify the trade-offs of the various transport protocols in a range of representative environments using DDS. }, URL = {http://www.truststc.org/pubs/675.html} }
Posted by Joe Hoffert on 29 Mar 2010.
Groups: trust
For additional information, see the
Publications FAQ or
contact webmaster at www truststc org.
Notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright.