Period integrals and Rankin-Selberg \(L\)-functions on \(\mathrm{GL}(n)\) (Q1938487): Difference between revisions

The paper under review deals with Rankin-Selberg \(L\)-functions attached to Hecke-Maaß forms on \(\mathrm{GL}(n)\). The first result can be stated as follows. \textit{Theorem 1:} Suppose \(f\in S_k\) is a holomorphic Hecke eigenform of even weight \(k\) for the group \(\mathrm{SL}_2({\mathbb Z})\). Let \(B_k\) be the orthogonal Hecke basis of \(S_k\). Then one has, for any \(\varepsilon>0\): \[ \sum_{g\in B_k}\mid L(f\times\overline{g},1/2)|^2\ll_\varepsilon k^{1+\varepsilon}. \] The second theorem concerns the more general case of the group \(\mathrm{SL}_n({\mathbb Z})\), \(n\geq 2\). More precsiely, let \(f\) be a tempered spherical Hecke-Maaß form for \(\mathrm{SL}_n({\mathbb Z})\) with spectral parameters \(\nu=(\nu_1,\cdots,\nu_{n-1})\in i{\mathbb R}^{n-1}\). Define the measure \[ d_{\mathrm{spec}}\nu=\Pi_{1\leq j\leq k\leq n-1}G(n(\nu_1+\cdots\nu_{n-1}))d\nu, \] where \(G(ix)=|\frac{\Gamma_{\mathbb R}(1+ix)}{\Gamma_{\mathbb R}(ix)}|^2\), \(\Gamma_{\mathbb R}(s)=\pi^{-s/2}\Gamma(s/2)\), and \(\Gamma\) is the Gamma-function. The second theorem can be stated as follows. \textit{Theorem 2:} \[ \sum_{\|\mu-\nu\|\leq 1}\mid L(f\times\overline{g},1/2)\mid^2\ll \big(\int_{\|\mu-\nu\|\leq 1}d_{\mathrm{spec}}\mu\big)^{1+\varepsilon} \] where the sum runs over an orthogonal basis of tempered Hecke-Maaß cusp forms \(g\) for \(\mathrm{SL}_n({\mathbb Z})\) with spectral parameter \(\mu\in i{\mathbb R}^{n-1}\) such that \(\|\mu-\nu\|\leq 1\). When \(n=2\) the above estimate reads as follows: \[ \sum_{\|\mu-\nu\|\leq 1}|L(f\times\overline{g},1/2)|^2\ll |\nu|^{1+\varepsilon}. \] In fact, if \(h\) is a fixed Hecke-Maaß cusp form for \(\mathrm{SL}_2({\mathbb Z})\) then one has (Theorem 3): \[ \sum_{\|\mu-\nu\|\leq 1}| L(f\times\overline{g}\times h,1/2)|^2 \ll_h|\nu|^{1+\varepsilon}. \] Moreover, these estimates are checked to be as strong as the Lindelöf hypothesis on average. The proofs are based on a geometric approach to \(L\)-functions, via periods of automorphic forms; avoiding trace formulae and explicit knowledge of Fourier coefficients.