Exponential sums nondegenerate relative to a lattice (Q969256)

The authors improve some of their important earlier work [e.g. \textit{A. Adolphson, S. Sperber}, Ann. Math. (2) 130, No. 2, 367--406 (1989; Zbl 0723.14017)], on estimates for exponential sums over the \(n\) dimensional torus over a fixed finite field or on properties the \(L\)-functions of families of such sums. Let \({\mathbb F}_q\) be a finite field with \(q\) elements and characteristic \(p\), and, for a subset \(J\) of \({\mathbb Z}^n\), \(\underline{x}=(x_1,\ldots,x_n)\), let \(f=\sum_{j\in J}a_j\underline{x}^j\in {\mathbb F}_q[x_1^{\pm 1},\ldots,x_n^{\pm 1}]\) (the symbol \(\underline{x}^j\) having an obvious meaning for \(j\in J\)). Denote further by \({\mathbb Z}\langle J\rangle\) (resp. \({\mathbb R}\langle J\rangle\)) the subgroup of \({\mathbb Z}^n\) (resp. \({\mathbb R}^n\)) generated by \(J\). The basis of the analysis relies on the fact that if we write \(f=g(x_1^{d_1},\ldots,x_k^{d_k})\) (\(k\) being the rank of \({\mathbb Z}\langle J\rangle\)) and \(d_i=p^{b_i}e_i\), \(p\nmid e_i\), the exponential sums associated to \(f\) and \(g(x_1^{e_1},\ldots,x_k^{e_k})\) are identical. The authors obtain improved versions of earlier results of theirs under the condition that the prime-to-\(p\) saturation of \({\mathbb Z}\langle J\rangle\): \[ M\langle J\rangle:=\{u\in {\mathbb Z}^n: \exists k, \,(k,p)=1 \text{ and }ku\in{\mathbb Z}\langle J\rangle\} \] is a proper subgroup of \({\mathbb Z}^n\cap {\mathbb R}\langle J\rangle\). The combined use of the set \(M\langle J\rangle\) and of the notion of non-degeneracy relative to a lattice introduced by the authors yields various results: they get a new way of computing the degree of the \(L\)-function of certain families of exponential sums considered by [\textit{N. M. Katz}, Moments, monodromy, and perversity. A Diophantine perspective. Annals of Mathematics Studies 159. Princeton, NJ: Princeton University Press (2005; Zbl 1079.14025)], the degree of the numerator of the zeta function of the Dwork family of hypersurfaces (also considered by \textit{A. Rojas-Leon} and \textit{D. Wan} [Commun. Number Theory Phys. 1, No. 3, 539--578 (2007; Zbl 1179.14024)]) and [\textit{N. M. Katz}, Algebra, arithmetic, and geometry. In honor of Y. I. Manin on the occasion of his 70th birthday. Vol. II. Boston, MA: Birkhäuser. Prog. Math. 270, 89--126 (2009; Zbl 1195.14015)], as well as a multivariate version of the Chevalley--Warning Theorem.