On additive complement of a finite set (Q2637185): Difference between revisions

From the text: Let \(\mathbb N\) denote the set of positive integers and let \(\mathcal A\subseteq\mathbb N\) and \(\mathcal B\subseteq\mathbb N\) be finite or infinite sets. Let \(R_{\mathcal A+\mathcal B}(n)\) denote the number of solutions of the equation \(a+b=n\), \(a\in\mathcal A\), \(b\in\mathcal B\). We put \[ A(n) =\sum_{_{\substack{ a\leq n\\ a\in\mathcal A}}} 1\quad\text{and}\quad B(n) =\sum_{_{\substack{ b\leq n\\ b\in\mathcal B}}} 1, \] respectively. We say a set \(\mathcal B\subseteq\mathbb N\) is an additive complement of the set \(\mathcal A\subseteq\mathbb N\) if every sufficiently large \(n\in\mathbb N\) can be represented in the form \(a+b=n\), \(a\in\mathcal A\), \(b\in\mathcal B\), i.e., \(R_{\mathcal A+\mathcal B}(n)\geq 1\) for \(n\geq n_0\). Additive complement is an important concept in additive number theory, in the past few decades it was studied by many authors. In [Acta Math. Hung. 64, No. 3, 237--245 (1994; Zbl 0816.11013)] \textit{A. Sárközy} and \textit{E. Szemerédi} proved a conjecture of \textit{L. Danzer} [J. Reine Angew. Math. 214/215, 392--394 (1964; Zbl 0126.27902)], namely they proved that for infinite additive complements \(\mathcal A\) and \(\mathcal B\) if \(\limsup_{x\to+\infty}\frac{A(x)B(x)}{x}\leq 1\), then \(\liminf_{x\to +\infty}(A(x)B(x)-x) = +\infty\). In [Proc. Am. Math. Soc. 138, No. 6, 1923--1927 (2010); 139, No. 3, 881--883 (2011; Zbl 1202.11010)] \textit{Y.-G. Chen} and \textit{J.-H. Fang} improved this result. In [Discrete Math. 313, No. 5, 595--598 (2013; Zbl 1311.11014)] they studied the case when \(\mathcal A\) is a finite set. In this case the situation is different from the infinite case. In this paper the authors prove a conjecture of Chen and Fang (Conjecture 1.5. in [Discrete Math. (loc. cit.)]), when the number of elements of the set \(\mathcal A\) is prime: Theorem 1. Let \(p\) be a positive prime and \(\mathcal A\) is a set of nonnegative integers with \(|\mathcal A| = p\). If \(\mathcal A\) is not of the form \[ \mathcal A = \{a + ip^s + k_ip^s+1 : i = 0, \ldots, p-1\}, \tag{1} \] where \(a > 0\), \(s\geq 0\) and \(k_i\) are integers, then, for any additive complement \(\mathcal B\) of \(\mathcal A\), we have \[ \lim_{x\to+\infty}(A(x)B(x)-x) = +\infty. \tag{2} \] In the case when the number of elements of \(\mathcal A\) is a composite number, we disprove the Conjecture 1.5. in [loc. cit.]: Theorem 2. For any composite number \(n > 0\), there exists a set \(\mathcal A\) and a set \(\mathcal B\) such that \(|\mathcal A| = n\), \(\mathcal B\) is an additive complement of \(\mathcal A\) and \(\mathcal A\) is not of the form \(\mathcal A = \{a + in^s + k_in^s+1 : i = 0, \ldots, n-1\}\), where \(s\geq 0\), \(a > 0\), and \(k_i\) are integers, and \[ A(x)B(x)-x = O(1). \] The authors also give a short survey about the algebraic concepts which play a crucial role in the proof of Theorem 1.