Random Walks and the Best Meeting Time for Trees
Andrew Beveridge, Ari Holcombe Pomerance
TL;DR
This work analyzes the best meeting time $T_{\mathrm{bestmeet}}(G)=\min_{v}\sum_{u}\pi_u H(u,v)$ for random walks on trees with order $n$ and diameter $d$, introducing the joining time $J(v)=2|E|H(\pi,v)$ and the barycenter as a centrality notion. The authors show that among such trees the balanced lever $L_{n,d}$ uniquely minimizes $T_{\mathrm{bestmeet}}$, while the balanced double broom $D_{n,d}$ uniquely maximizes it, deriving explicit parity-dependent formulas for these extremal values. They also provide global bounds for all trees of order $n$, showing $T_{\mathrm{bestmeet}}(S_n)=\tfrac{1}{2}$ as the minimum, and giving maximizing structures as the path $P_n$ when $n$ is even and the broom $B_{n,n-2}$ (rather than the path) when $n$ is odd and large enough. The results identify how diameter-constrained tree shapes influence centrality-like measures of random walks and offer precise, exploitable formulas for extremal meeting times.
Abstract
We consider random walks on a tree $G=(V,E)$ with stationary distribution $π_v = \mathrm{deg}(v)/2|E|$ for $v \in V$. Let the hitting time $H(v,w)$ denote the expected number of steps required for the random walk started at vertex $v$ to reach vertex $w$. We characterize the extremal tree structures for the best meeting time $T_{\mathrm{bestmeet}}(G) = \min_{w \in V} \sum_{v \in V} π_v H(v,w)$ for trees of order $n$ with diameter $d$. The best meeting time is maximized by the balanced double broom graph, and it is minimized by the balanced lever graph.
