The Adaptive Systems Project is developing principles, methods, algorithms,
and system support for the online adaptation of applications and system
components, to address the needs of embedded/real-time, high performance,
and interactive applications.
Current research efforts are funded by federal and industry sponors.
infrastructures for online system/program adaptation, such as the FARA
infrastructure of runtime adaptation of complex distributed and parallel
applications subject to constraints in their timing behavior, and the Dionisys
infrastructure of adaptation of multiple application and system components
cooperating in the execution of a single complex application,
middleware and software abstractions enabling the runtime control of applications,
application components, and system-level services,
experimentation with specific adaptation techniques, adaptive applications,
software and hardware components, and abstractions supporting runtime adaptation,
adaptable real-time programs, real-time benchmarks, and real-time protocols:
real-time distributed and parallel programs can exhibit highly dynamic
execution behavior. We have developed several distributed and parallel
benchmark programs operating on multiple target platforms. These
programs include synthetic applications derived from distributed multimedia
and from parallel sensor processing applications, distributed virtual environments,
and distributed games and simulations. Jointly, they provide a basis for
research on configurable and adaptable distributed programs and communications.
Our research addresses the development of on-line adaptation algorithms
and methods for parallel, distributed, and real-time applications. Toward
this end, negotiation protocols and QoS formulations targeting real-time
applications are being investigated, and a framework for dynamic resource
management is being constructed. This project also concerns the evaluation
of on-line configurable communication protocols, both with respect to their
processing requirements (CPU requirements on SMP platforms) and their networking
needs (QoS requirements), jointly with networking faculty.
Robin Kravets, Ken Calvert, and Karsten Schwan, Payoff
Adaptation of Communication for Distributed Interactive Applications,
to appear in Journal on High Speed Networking: Special Issue on Multimedia
Bodhi Mukherjee, Dilma Silva, Karsten Schwan, and Ahmed Gheith,
KTK: Kernel Support for Configurable Objects and Invocations, Distributed
Systems Engineering Journal, IEE, UK, Vol.1, 1995.
Ahmed Gheith and Karsten Schwan, CHAOS-Arc -- Kernel Support for Atomic
Transactions in Real-Time Applications, ACM Transactions on Computer
Systems, April 1993. Also reprinted in Advances in Real-time Systems, Prentice
Hall, editor: S. Son, 1994.
Thomas E. Bihari and Karsten Schwan, The Dynamic Adaptation of Real-Time
Software, ACM Transactions on Computer Systems, May 1991.
Robin Kravets, Ken Calvert, and Karsten Schwan, Payoff-Based
Communication Adaptation based on Network Service Availability, IEEE
Multimedia Systems '98, Aug. 1998.
Daniela Ivan-Rosu and Karsten Schwan, FARA--
A Framework for Adaptive Resource Allocation in Complex Real-Time Systems,
IEEE Real-Time Technology and Applications Symposium, June 1998.
Dilma Silva, Karsten Schwan, and Greg Eisenhauer, Configurable
Distributed Retrieval of Scientific Data, International Conference
on Configurable Distributed Systems, May 1998.
Vernard Martin and Karsten Schwan, ILI:
An Adaptive Infrastructure For Dynamic Interactive Distributed Applications,
International Conference on Configurable Distributed Systems, May 1998.
Daniela Ivan-Rosu, Karsten Schwan, Sudhakar Yalamanchili, and Rakesh Jha,
Adaptive Resource Allocation for Complex, Real-time Applications, Real-Time
Systems Symposium, San Francisco, IEEE, Dec. 1997.
Phyllis Schneck, Ellen Zegura, and Karsten Schwan, DRRM: Dynamic Resource
Reservation Manager, Proceedings of the IC3N 96, IEEE, October 1996.
Daniela Ivan-Rosu and Karsten Schwan, Improving
Protocol Performance by Dynamic Control of Communication Resources,
Second IEEE International Conference on Engineering of Complex Computer
Systems, Montreal, October 1996, outstanding paper award.
Bodhi Mukherjee and Karsten Schwan, Improving Performance by Use of
Adaptive Objects: Experimentation with a Configurable Multiprocessor Threads
Package, 2nd International Symposium on High Performance Distributed
Computing (HPDC-2), Spokane, Washington, July 1993.
Thomas E. Bihari and Karsten Schwan, A Comparison of Four Adaptation
Algorithms for Increasing the Reliability of Real-Time Software, Ninth
Real-Time System Symposium, Huntsville, AL, Dec. 1988.
Richard West and Karsten Schwan, Interactors:
Capturing QoS and Resource Requirements between Multiple Cooperating Objects,
Fourth IEEE Real-Time Technology and Applications Symposium, Work-In-Progress,
IEEE, June 1998.
P. Schneck and K. Schwan, Dynamic Allocation of Security Resources to
Client-Server Applications, IEEE Workshop on Dependable and Real-Time
E-Commerce Systems, May 1998.
P. Schneck and K. Schwan, Dynamic Authentication for High-Performance
Networked Applications, 1998 Sixth IEEE/IFIP International Workshop
on Quality of Service, May 1998.
D. Silva and K. Schwan, Achieving High Performance through Flexible
Software, Proceedings of the XXIV Brazilian Software and Hardware Seminars
(SEMISH), Aug. 1997.
Projects and Links
[Georgia Institute of
Technology][College of Computing]
Last modified: September 16, 1998.
Page maintained by Rich West, email@example.com