×

A proof of Boesch’s conjecture. (English) Zbl 0826.90060

Summary: A well-known model for network reliability studies consists of an undirected graph with perfectly reliable nodes and equal and independent edge failure probabilities. The measure of reliability is defined to be the probability that a graph is connected. A well-defined synthesis problem is to find a graph that maximizes the probability given the number of nodes \(n\), the number of edges \(e\), and edge failure rate \(\rho\). The problem of a uniformly optimally reliable graph is to find a graph that is optimal for all possible \(\rho\). The graph studies are attracting more and more attention. F. T. Boesch, X. Li and C. Suffel [Networks 21, No. 2, 181-194 (1991; Zbl 0721.90038)] verified the existence of uniformly optimally reliable graphs when \(e= n\), \(n+1\), \(n+2\) and given a conjecture for \(e= n+ 3\). In this work, a proof of the conjecture is given and it is proved that any min-\(m_i\) graph is 2-connected.

MSC:

90B25 Reliability, availability, maintenance, inspection in operations research
90C35 Programming involving graphs or networks

Citations:

Zbl 0721.90038
Full Text: DOI

References:

[1] Bauer, Networks 15 pp 257– (1985)
[2] Bauer, IEEE Trans. Circ. Syst. 34 pp 1579– (1989)
[3] Boesch, J. Graph Theory 10 pp 339– (1986)
[4] Boesch, Networks 21 pp 181– (1991)
[5] and , Graph Theory with Applications. Flsevier North-Holland, New York (1976). · Zbl 1226.05083 · doi:10.1007/978-1-349-03521-2
[6] Some properties of t-optimal graphs. ICCAS (1989) 736–739.
[7] On the synthesis of reliable networks. PhD Thesis, Stevens Institute of Technology (1986) 61.
[8] Li, J. Comput. 9 pp 699– (1990)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.