Non-existence of some generalized bent functions. (Q1566013)

The authors prove some new nonexistence theorems of generalized bent functions going on with the method used in [\((*)\) \textit{K. Feng}, Sci. China, Ser. A 44, 562--570 (2001; Zbl 1003.11060)]. They begin by defining an \([n, q]\) generalized bent function (GBF) as follows: \(f: \mathbb{Z}^n_q\to \mathbb{Z}_q\), \(|\sum_{x\in \mathbb{Z}^n_q} \zeta^{f(x)- x\cdot y}_q|= q^{{n\over 2}}\), where \(q> 2\) and \(n\) are positive integers, \(\mathbb{Z}_q\) is the ring of integers modulo \(q\) and \(\zeta_q\) a \(q\)th root of unity in \(\mathbb{C}\). They focus their results on this family of parameters: \(n\) odd and \(q= 2N\), \(2\nmid N\geq 3\) with \(N= p_1p_2\) and \(p_1\), \(p_2\) distinct odd primes. They recall a technical lemma from [\((*)\)] useful in the next parts of the paper. The authors treat about three cases depending on the values of parameters \(p_1\) and \(p_2\). For each case, the related scheme of extension fields is clearly described, a theorem and its corollary are proved, and a short example illustrates the theorical result. The goal of this work is cleanly reached with some new results on the nonexistence of \([n, q]\) GBF for \(n\) odd and \(q= 2N\), \(2\nmid N\), \(N\geq 3\), \(N= p_1p_2\) with \(p_1\), \(p_2\) distinct odd prime numbers. Interesting properties of the decomposition of rational fields and results on Diophantine equations are brilliantly used to obtain their theorems. Citations are correctely used, but results are not well presented. The outline of the paper is not very clear. It's a bit difficult to distinguish the different parts of the paper: Section 1: Introduction; Section 2: \(p_1\equiv 2^\lambda+ 1\pmod{2^{\lambda+1}}\), \(\lambda\geq 3\), \(p_2\equiv 7\pmod 8\), \(({p_1\over p_2})= 1\), \(({p_2\over p_1})_4\neq 1\), \(({2\over p_1})_4\neq 1\); Section 3: \(p_1\equiv 5\pmod 8\), \(p_2\equiv 7\pmod 8\), \(({p_1\over p_2})= 1\), \(({p_2\over p_1})_4\neq 1\); Section 4: \(p_1\equiv 5\pmod 8\), \(p_2\equiv 3\pmod 8\),\(({p_1\over p_2})= 1\), \(({p_2\over p_1})_4\neq 1\). It would be more clever to present for example section 4 in this way: Section 4 -- case No. 3: \(p_1\equiv 5\pmod 8\), \(p_2\equiv 3\pmod 8\), \(({p_1\over p_2})= 1\), \(({p_2\over p_1})_4\neq 1\). Moreover, a conclusion would have been appreciated with a little sum up and some further open problems on other possible cases related to a eventual classification. See also the second part of the authors' paper (with \textit{Z. Ma}) in Sci. China, Ser. A 45, No. 6, 721--730 (2002).