Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Graceful fault tolerance in large networks of microcomputers

Thesis/Dissertation ·
OSTI ID:5594631
This work considers the problem of fault diagnosis in a network of distributed multicomputers, and a strategy for repeated reconfiguration is presented in detail to help improve the degree of fault tolerance. The overall system diagnosability is shown to be enhanced further by constructing a large network with small well-known graphs as its basis and then applying reconfiguration techniques locally in various system partitions and exchanging diagnostic information globally. A detailed description of this new attractive approach is presented along with the diagnostic algorithm suitable for large networks of microcomputers in VLSI based distributed systems. A systematic procedure for defining near-optimal fault-tolerance graph theoretic networks is investigated which is well suited for multicomputer structures. A distributed algorithm along with a new system diagnostic theory is proposed.
Research Organization:
Western State Coll. of Colorado, Gunnison (USA)
OSTI ID:
5594631
Country of Publication:
United States
Language:
English

Similar Records

Partitioning of large multicomputer systems for efficient fault diagnosis
Conference · Thu Dec 31 23:00:00 EST 1981 · OSTI ID:5000331

A class of fault-tolerant multiprocessor networks
Journal Article · Fri Mar 31 23:00:00 EST 1989 · IEEE (Institute of Electrical and Electronics Engineers) Transactions on Reliability; (USA) · OSTI ID:5242595

Fault tolerance in multistage interconnection network-based multicomputer systems
Thesis/Dissertation · Wed Dec 31 23:00:00 EST 1986 · OSTI ID:5705671