Exponential sums and stationary phase. I (Q4485873)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1461543
Language Label Description Also known as
default for all languages
No label defined
    English
    Exponential sums and stationary phase. I
    scientific article; zbMATH DE number 1461543

      Statements

      20 June 2000
      0 references
      stationary phase
      0 references
      Mellin transform
      0 references
      exponential sums
      0 references
      Gauss sums
      0 references
      Exponential sums and stationary phase. I (English)
      0 references
      The classical method of stationary phase yields asymptotic results for integrals of the form \(\int_M a(y)e^{2\pi ikf(y)} dy\), where \(M\) is a Riemannian manifold. Under suitable conditions, the leading term has the shape \(\sum_{s \text{in} C_f} a(s)e^{2\pi i\text{sgn}H(s)/8} {e^{2\pi ikf(s)}\over\sqrt{|\det H(s)|}}\), where \(C_f=\{s \text{in} M:(df)_s=0\}\) is the set of critical points of \(f\) and \(H(s)=(d^2f)_s\) is the Hessian matrix. Igusa developed a \(p\)-adic version of the method for local Mellin transforms. Consider a polynomial \(f\) in \({\mathbb{Z}}_p[x_1,\ldots,x_n]\) and a point \(\mu_1=(y_1,\ldots,y_n)\) in \({\mathbb{Z}}_p^n\). Set \(f_\mu(x)=p^{-e_\mu}f(px+\mu)\) where \(e_\mu\) is the minimum order of \(p\) in the expansion of \(f(px+\mu)\). Let \(V_f(A)\) denote the set of \(A\)-valued points, \(C_f(A)\) the \(A\)-valued critical set, \(\nu(f)=\#\{\mu:\mu\) is not in \(V_f({\mathbb{F}}_p)\}\) and \(\sigma(f)=\#\{\mu:\mu\) is in \(C_f({\mathbb{F}}_p)\}\). Again, under suitable conditions, Igusa's formula is \[ \int_X|f(x)|_p^s dx=\left\{\nu(f) +\sigma(f){(1-p^{-1})p^{-s}\over 1-p^{-1}p^{-s}} +\sum_{\mu \text{in} C_f(X)} q^{-e_\mu s} \int_{{\mathbb{Z}}_p^n} |f_\mu(x)|_p^s dx\right\}q^{-n}. \] Such formulae have applications to the determination of exponential sums. A classical example is the evaluation of the quadratic Gauss sum \[ \sum_{s\bmod n}e^{2\pi is^2/n}={1+(-i)^n\over 1-i}\sqrt n. \] The author sketches the generalised Gauss sums of Lamprecht and Dwork along with their evaluations. Another example is \[ \sum_{s\bmod p^l}e^{2\pi if(s)/p^l}=\sum_j a(j)(\sqrt p)^{w(j)} \] where \(f\) is a polynomial and the sum on the right has \(|a(j)|=1\) and the weight exponents \(w(j)\) are integers in the range \(\{0,1,\ldots,l\}\) arising from the critical points of \(f\). Igusa's theory gives generalisations to polynomials in several variables over finite rings. The present paper is a survey of these ideas and a further paper is foreshadowed to take up the explicit determination of exponential sums in more detail. For the entire collection see [Zbl 0930.00089].
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references