The characteristic polynomial of a random unitary matrix: a probabilistic approach (Q953965)

The characteristic polynomial \(Z_N\) of an \(\mathbb N \times\mathbb N\) random unitary matrix drawn from the unitary group \(U(N)\) with the Haar measure \(\mu_{U(N)}\) is studied. The convergence in law of\break \(\log Z_N/\sqrt{(1/2)\log N}\) to a standard complex Gaussian law is derived as a consequence of classical central limit theorem. Using a recursive construction for the Haar measure on \(U(N)\), the equality in law for the distribution of the characteristic polynomial as a product of independent random variables is obtained. From this equality the Mellin-Fourier transform of the random polynomial \(Z_N\), first obtained by \textit{J. P. Keating} and \textit{N. C. Snaith} [Commun. Math. Phys. 214, No.~1, 57--89 (2000; Zbl 1051.11048)], is derived without using Selberg's integrals or the Weyl density. Then, the joint law of \((\Re\log Z_N, \Im\log Z_N)\), writing each component as a sum of independent random variables, is deduced. These two representations for \(Z_N\) are used to a new proof of the Keating-Snaith limit theorem. Estimates on the rate of convergence in the central limit theorem are also given and it is shown how strong limit theorems such as the iterated logarithm can be deduced from the presented representations. Some results are given for the characteristic polynomial of other than the unitary group and namely for the group \(\text{ SO}(2N)\) which plays a similar role to \(U(N)\) for other families of \(L\)-functions.