A new framework for identifying most reliable graphs and a correction to the $K_{3,3}$-theorem
Abstract
Given a multigraph , the all-terminal reliability is the probability that remains connected under percolation with parameter . Fixing the number of vertices and edges , we investigate which graphs maximize -- such graphs are called optimal -- paying particular attention to uniqueness and to whether the answer depends upon . We generalize the concept of a distillation and build a framework with which we identify all optimal graphs for which . These graphs are uniformly optimal in . Most have been previously identified, but with serious problems, especially when . We obtain partial results for . For , the optimal graphs were incorrectly identified by Wang in 1994, in the infinite number of cases where and . This erroneous result concerns subdivisions of and has been cited extensively, without any mistake being detected. While the optimal graphs were correctly described for other , the proof is fundamentally flawed. Our proof of the rectified statement is self-contained. For , the optimal graphs were recently shown to depend upon for infinitely many . We find a new such set of -values, which gives a different perspective on why this phenomenon occurs and leads us to conjecture that uniformly optimal graphs exist only for finitely many . However, for , we conjecture that there are again infinitely many uniformly optimal graphs.
Cite
@article{arxiv.2407.20217,
title = {A new framework for identifying most reliable graphs and a correction to the $K_{3,3}$-theorem},
author = {Lorents F. Landgren and Jeffrey E. Steif},
journal= {arXiv preprint arXiv:2407.20217},
year = {2024}
}
Comments
44 pages, 26 figures. Some shortened arguments, clarifications, copyediting, two new figures. Comments welcome