Arithmetical progressions formed by \(k\) different Lehmer pseudoprimes (Q1896380): Difference between revisions

A Lehmer sequence of integers is defined by \(U_n = (\alpha^n - \beta^n) /(\alpha - \beta)\) for \(n\) odd and \(U_n = (\alpha^n- \beta^n)/(\alpha^2 - \beta^2)\) for even \(n\), where \(\alpha\) and \(\beta\) are the distinct roots of the polynomial \(z^2 - \sqrt L z + Q\) and \(L\) \((> 0)\) and \(Q\) are fixed integers. We say a composite \(n\) is a Lehmer pseudoprime for the sequence \(U\) if \(U_{n - (LD/n)} \equiv 0 \pmod n\), where \(D = L - 4Q\) and \((LD/n)\) is the Jacobi symbol. For Lehmer sequences with \(D > 0\) the author proves: if \(p > 3\) and \(2p - 1\) are primes satisfying some conditions, then there are three numbers which are Lehmer pseudoprimes and form an arithmetic progression. In the case \(D > 0\) it is also shown that for every positive integer \(k\) Schinzel's hypothesis \(H\) implies the existence of infinitely many arithmetic progressions formed from \(k\) distinct Lehmer pseudoprimes.