Detail publikace

Evolutionary Design of Fault Tolerant Collective Communications

Originální název

Evolutionary Design of Fault Tolerant Collective Communications

Anglický název

Evolutionary Design of Fault Tolerant Collective Communications

Jazyk

en

Originální abstrakt

Scheduling of collective communications (CC) in interconnection networks possibly containing faulty links has been done with the use of the evolutionary techniques. Inter-node communication patterns scheduled in the minimum number of time slots have been obtained. The results show that evolutionary techniques often lead to ultimate scheduling of CC that reaches theoretical bounds on the number of steps. Analysis of fault tolerance by the same techniques revealed graceful CC performance degradation for a single link or node fault. Once the faulty region is located, CC can be re-scheduled during a recovery period.

Anglický abstrakt

Scheduling of collective communications (CC) in interconnection networks possibly containing faulty links has been done with the use of the evolutionary techniques. Inter-node communication patterns scheduled in the minimum number of time slots have been obtained. The results show that evolutionary techniques often lead to ultimate scheduling of CC that reaches theoretical bounds on the number of steps. Analysis of fault tolerance by the same techniques revealed graceful CC performance degradation for a single link or node fault. Once the faulty region is located, CC can be re-scheduled during a recovery period.

BibTex


@inproceedings{BUT30898,
  author="Jiří {Jaroš}",
  title="Evolutionary Design of Fault Tolerant Collective Communications",
  annote="Scheduling of collective communications (CC) in interconnection networks possibly
containing faulty links has been done with the use of the evolutionary
techniques. Inter-node communication patterns scheduled in the minimum number of
time slots have been obtained. The results show that evolutionary techniques
often lead to ultimate scheduling of CC that reaches theoretical bounds on the
number of steps. Analysis of fault tolerance by the same techniques revealed
graceful CC performance degradation for a single link or node fault. Once the
faulty region is located, CC can be re-scheduled during a recovery period.",
  address="Springer Verlag",
  booktitle="Evolvable Systems: From Biology to Hardware",
  chapter="30898",
  edition="Lecture Notes in Computer Science",
  howpublished="online",
  institution="Springer Verlag",
  journal="Lecture Notes in Computer Science (IF 0,513)",
  year="2008",
  month="september",
  pages="261--272",
  publisher="Springer Verlag",
  type="conference paper"
}