Mean-regime conjecture for the maximum and extremal process of two-speed branching random walk

About 1 year old · traced to

Suppose that L1\mathcal{L}_1 and L2\mathcal{L}_2 are non-lattice. Assume that there exist θ1∗,θ2∗>0\theta_1^*,\theta_2^*>0 such that L1\mathcal{L}_1 and L2\mathcal{L}_2 satisfy assumptions (6), (7), and (8), with θ1∗=θ2∗\theta_1^*=\theta_2^*. Let θ:=θ1∗\theta:=\theta_1^* and

mn:=κ1′(θ)⌊tn⌋+κ2′(θ)(n−⌊tn⌋)−32θlog⁡n.m_n:=\kappa_1'(\theta)\lfloor tn\rfloor+\kappa_2'(\theta)(n-\lfloor tn\rfloor)-\frac{3}{2\theta}\log n.

Mean-regime conjecture. There exists a constant Cm>0C_{\mathrm{m}}>0 such that, for every y∈Ry\in\mathbb{R},

lim⁡n→∞P(Mn(n)−mn≤y)=E(e−CmZ(1)e−θy),\lim_{n\to\infty}\mathbb{P}(M_n^{(n)}-m_n\leq y)=\mathbb{E}\left(e^{-C_{\mathrm{m}}Z^{(1)}e^{-\theta y}}\right),

and the extremal process

En:=∑∣u∣=nδV(n)(u)−mn\mathcal{E}_n:=\sum_{|u|=n}\delta_{V^{(n)}(u)-m_n}

converges in law to a randomly shifted decorated Poisson point process E\mathcal{E} with random intensity measure CmZ(1)θe−θydyC_{\mathrm{m}}Z^{(1)}\theta e^{-\theta y}dy and decoration D(2)\mathcal{D}^{(2)}. Here D(2)\mathcal{D}^{(2)} is the limit in law of

∑∣u∣=nδV2(u)−max⁡∣u∣=nV2(u)\sum_{|u|=n}\delta_{V_2(u)-\max_{|u|=n}V_2(u)}

conditioned on {max⁡∣u∣=nV2(u)≥κ2′(θ)n}\{\max_{|u|=n}V_2(u)\geq\kappa_2'(\theta)n\}. The conjecture concerns the as-yet unstudied mean regime for branching Brownian motion; it predicts both the limiting law of the centered maximum and the limiting extremal process.

References

Primary source

Lianghui Luo, “The extremal process of two-speed branching random walk”, arXiv:2503.05994 (2025).

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.