Essentialness in additive bases (Q868906): Difference between revisions

A subset \(A\) of the positive integers is a base of order \(h\) if all sufficiently large positive integers are a sum of at most \(h\) elements of \(A\). For example, the primes are a base of order \(4\). Call a finite part \(P\) of a base \(A\) essential if \(A\backslash P\) is no longer a basis. For example, \(A=\{1\}\cup \{2n: n\geq 1\}\) is a base of order \(2\) for which \(P=\{1\}\) is essential. Clearly, if \(P\) is essential and \(P\subset P'\), then \(P'\) is also essential, so it makes sense to look at essential subsets of \(A\) which are minimal with respect to inclusion. In a preceding work on this topic [J. Reine Angew. Math. 539, 45--53 (2001; Zbl 1002.11011)], the first author and \textit{G. Grekos} showed that minimal essential subsets of \(A\) have order of magnitude at most \(\sqrt{{h}\over {\log h}}\) and this is sharp. In this paper, the authors find the best multiplicative constant namely \(30\sqrt{{\log 1564}\over {1564}}\) and show that this is sharp by exhibiting a base of order \(1564\) with a minimal essential subset of cardinality \(30\). The paper also contains some results on the structure of bases possessing minimal essential subsets. They show that, up to a finite subset, such sets are of the form \(aX+b\), where \(X\) is some subset of the natural numbers, and that the number of \(a\)'s for which \(A\) looks like above (where \(X\) and \(b\) are allowed to vary with \(a\)) is finite. Calling the largest such \(a\) the ``motif'' of \(A\), the authors show, among other things, that the cardinality of the largest minimal essential subset of \(A\) does not exceed the number of prime factors (counted with multiplicity) of the motif of \(A\). The paper concludes with two open problems.