Crossover-regime mixing-time conjecture for the weakly asymmetric simple exclusion process

Let bNb_N and kNk_N satisfy

limNbNNlogkN=β\lim_{N\to\infty}\frac{b_NN}{\log k_N}=\beta

for some β(0,)\beta\in(0,\infty), with limNkN/N=0\lim_{N\to\infty}k_N/N=0. Write TmixN,kN(ε)T_{\rm mix}^{N,k_N}(\varepsilon) for the mixing time of the weakly asymmetric simple exclusion process. Crossover-regime mixing-time conjecture. For every ε>0\varepsilon>0,

limNTmixN,kN(ε)logkNN2={2β+1β2,if β    1/2,(2+2β2β)2,if β    1/2.\lim_{N\to\infty}\frac{T_{\rm mix}^{N,k_N}(\varepsilon)\log k_N}{N^2}=\begin{cases}\frac{2}{\beta}+\frac{1}{\beta^2},&\text{if }\beta\;\leqslant\; 1/2,\left(\frac{\sqrt{2}+2\sqrt{\beta}}{2\beta}\right)^2,&\text{if }\beta\;\geqslant\; 1/2. \end{cases}

This gives a conjectural mixing-time asymptotic in the crossover regime bNlogkN/Nb_N\asymp\log k_N/N; the paper provides heuristic support, while the assertion remains unproved in the stated generality.

Sources & referencesView supporting material

Primary source

C. Labbé and H. Lacoin, “Mixing time and cutoff for the weakly asymmetric simple exclusion process”, arXiv:1805.12213 (2018).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.