Covering intervals with arithmetic progressions (Q2204100)

Let \(\mathcal{A}=\{\,A_1,A_2,\dots,A_k\,\}\) be a family of \(k\) arithmetic progressions \(A_i=a_i\mod{d_i}\). If a set of integers \(S\) is contained in the union \(\cup_iA_i\) we say that \(\mathcal{A}\) covers \(S\). And if \(\mathcal{A}\) covers all of \(\mathbb{Z}\), \(\mathcal{A}\) is said to be a covering system. The study of covering systems was initiated by \textit{P. Erdős} and, among other things, he conjectured [Mat. Lapok 13, 228--254 (1962; Zbl 0127.02202)] that if \(\mathcal{A}\) covers the set \([2^k]=\{1,2,\dots,2^k\}\), then, in actual fact, \(\mathcal{A}\) covers all of \(\mathbb{Z}\). It is easy to show that \(\mathcal{A}\) could cover the set \([2^k-1]\) without being a covering system.\par This conjecture was first proved by \textit{R. B. Crittenden} and \textit{C. L. Vanden Eynden} [Proc. Am. Math. Soc. 24, 475--481 (1970; Zbl 0192.39001)], and the paper under review gives a new proof of this result -- this new proof is surely from ``The Book''. In addition, the authors give the following two corollaries, the second of which was stated by Erdős as a separate conjecture.\par If \(\mathcal{A}\) is a covering system but no proper subset of \(\mathcal{A}\) is a covering system, then all of its moduli \(d_i\le2^{k-1}\).\par If moduli of \(\mathcal{A}\) satisfy \(\sum_{i=1}^k1/d_i<1\), then no set of \(2^k\) consecutive numbers is covered by \(\mathcal{A}\).