Next-order asymptotic expansion of the optimal Hardy weight in dimensions d ≥ 3

Establish the conjectured next-order asymptotic expansion of the optimal Hardy weight associated with the simple random walk on \(\mathbb{Z}^d\), \(d\geq 3\), namely \(w(x)=\frac{(d-2)^2}{8d|x|^2}+\frac{Q(\omega^x)}{|x|^4}+\mathcal{O}(|x|^{-6})\), where \(\omega^x=x/|x|\) and \(Q(z)=\frac{(d-2)^2}{16d}\left(\frac{5(d+2)^2}{8}\sum_{j=1}^d z_j^4-(2d+3)\right)\).

Background

For transition weights on Zd\mathbb{Z}^d with d3d\geq 3, the paper proves that the Hardy weight constructed from the square root of the Green function has leading asymptotic behavior wμ(x)=(d2)2σμ28xμ2+O(x4)w_{\mu}(x)=\frac{(d-2)^2\sigma_\mu^2}{8|x|_\mu^2}+\mathcal{O}(|x|^{-4}) under a suitable moment condition.

In the special case of the simple random walk, the authors note that the explicit higher-order term in the Green-function expansion appears sufficient to derive a further term in the Hardy-weight expansion. They explicitly label the displayed formula as a conjecture, leaving the proposed x4|x|^{-4} correction and its angular dependence unresolved.

References

For the simple random walk, with the explicit formula of the higher-order term (see the Remark following Theorem~\ref{thm:UchiyamaFukai2}) it should be possible to derive the asymptotic expansion for w = \frac{\Delta G{1/2}}{G{1/2}} with one more higher order term. After consultation with AI we conjecture it to be

w_{\mu}(x) = \frac{(d-2)2}{8 d|x|2} + \frac{Q(\omega_ix)}{|x|4} + \mathcal{O}\big(|x|{-6}\big),

with Q given as

Q(z) = \frac{(d-2)2}{16d} \left( \frac{5(d+2)2}{8} \sum_{j=1}d z_j4 -(2d+3) \right).

Optimal Hardy Inequalities for Random Walks on $\mathbb{Z}^2$  (2608.24213 - Hake et al., 25 Aug 2026) in Remark following Proposition in Section “Expansion of \(w_{\mu}\) for \(d\geq 3\)”

It remains to determine exact pure-power constants in fixed dimensions $d\geq4$ and for nonlinear discrete $p$-energies.

The sharp discrete Hardy inequality on $\Z^3$  (2608.25262 - Alpay, 26 Aug 2026) in Remark ‘Further problems’