Product Bases for the Rationals (Q4944105)

Let \(Q^*\) denote the multiplicative group of positive rationals, \(\Gamma\) the subgroup generated by the members, \(a_n\), of a sequence of rationals, \(G\) the quotient group, \(Q^*/\Gamma\). For each positive integer \(t\), \(S_t\) denotes the members of \(\Gamma\) representable by a group product using at most \(t\) terms from the sequence \((a_n)\), repetition allowed. For a set of rationals, \(E\), define the lower density \[ d(E)= \liminf_{x\to\infty} x^{-1} \sum_{n\leq x, n\in E}1, \] the sum counting only positive integers. Then the authors prove the following: Theorem. The following propositions are equivalent: (i) \(G\) is finite and \(\Gamma= S_t\) for some \(t\). (ii) There is a \(k\) such that \(d(m^{-1}S_k)\geq c_1> 0\) uniformly for all positive integers \(m\). (iii) There is a \(w\) such that \(d(\gamma S_w)\geq c_2> 0\) uniformly for all positive rationals \(\gamma\). Moreover, \(t\) can be given explicitly in terms of \(k\) and \(c_1\), or \(w\) and \(c_2\). The proof involves methods of Elliott and Ruzsa. Combined with results from number theory, a careful variant of the argument given here shows that any rational representable as a product of shifted primes \(p+1\), has such a representation with exactly 19 primes [the authors, Ramanujan J. 2, 219-223 (1998; Zbl 0907.11032)] and moreover exactly 9 primes [the second author, J. Reine Angew. Math. 519, 59-71 (2000; Zbl 1044.11088)].