A new framework for identifying most reliable graphs and a correction to the $K_{3,3}$-theorem
Lorents F. Landgren,Jeffrey E. Steif
2024-07-30
Abstract:Given a finite connected multigraph $G=(V,E)$, the all-terminal reliability function $R(G,p)$ is the probability that $G$ remains connected under percolation with parameter $p$. We develop new results concerning the question of which (multi)graphs~$G$ maximize $R(G,p)$ for a given number of vertices $n$, edges $m$ and a given parameter $p$ -- such graphs are called optimally reliable graphs -- paying particular attention to uniqueness and to whether the answer depends upon $p$.
We generalize the concept of a distillation, and build a framework which we use to identify all optimally reliable graphs for which $m-n\in\{1,2,3\}$. These graphs are uniformly optimal in $p$. Many of them have been previously identified, but there are serious problems with these treatments, especially for the case $m-n=3$. We also obtain some partial results for $m-n\in\{4,5\}$.
For $m-n=3$, the optimal graphs were incorrectly identified by Wang in 1994, in the infinite number of cases where $m\equiv 5 \pmod{9}$ and $m\geq 14$. This erroneous result, which concerns subdivisions of $K_{3,3}$, has been cited extensively and the mistake has not been detected. Furthermore, while the optimal graphs were correctly identified for other values of $m$, the proof is incorrect, as it also uses a claim for which we give a counterexample. Our proof of the rectified statement is self-contained.
Regarding $m-n=4$, it was recently shown by Romero and Safe that the optimal graph depends upon $p$ for infinitely many $m$. We find a new such set of $m$-values, which gives a different perspective on why this phenomenon occurs. This leads us to conjecture that there are only finitely many $m$ for which there are uniformly optimal graphs when $m-n=4$, but that, for $m-n=5$, there are again infinitely many uniformly optimal graphs. We hope that our general results and methods will be useful for continued investigations.
Combinatorics,Probability