Single-root Fourier-eigenfunction convergence conjecture

About 10 years old · traced to

For ε∈{0,1}\varepsilon\in\{0,1\}, let

gn,kε(x)=rn,kε(∣x∣2)e−π∣x∣2,g^\varepsilon_{n,k}(x)=r^\varepsilon_{n,k}(|x|^2)e^{-\pi|x|^2},

where rn,kεr^\varepsilon_{n,k} has degree at most 2k+ε2k+\varepsilon, is nonzero, vanishes at 2,4,…,2k2,4,\ldots,2k, and is a linear combination of the relevant radial Fourier-eigenpolynomials, so that g^n,kε=(−1)εgn,kε\widehat{g}^\varepsilon_{n,k}=(-1)^\varepsilon g^\varepsilon_{n,k}. Single-root Fourier-eigenfunction convergence conjecture. With nn and ε\varepsilon fixed, as k→∞k\to\infty, suitably normalized gn,kεg^\varepsilon_{n,k} converges to a nonzero Fourier eigenfunction gnεg^\varepsilon_n vanishing at every radius 2j\sqrt{2j}, and the convergence is uniform on compact subsets of C\mathbb C when the functions are viewed as entire functions of ∣x∣|x|. Uniform convergence would imply that the limiting eigenfunctions are entire; the source states the convergence conjecturally and gives no general proof.

References

Primary source

Henry Cohn and Stephen D. Miller, “Some properties of optimal functions for sphere packing in dimensions 8 and 24”, arXiv:1603.04759 (2016).

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.