The proof of a conjecture of additive number theory (Q1609535)

From MaRDI portal
scientific article
Language Label Description Also known as
English
The proof of a conjecture of additive number theory
scientific article

    Statements

    The proof of a conjecture of additive number theory (English)
    0 references
    0 references
    15 August 2002
    0 references
    For any positive integer \(n\) let \(\Omega(n)\) stand for the number of prime divisors of \(n\) counted with multiplicities. The paper deals with the following type of questions: Is it true that all positive integers \(n\), with a few small exceptions, can be written as a sum \(n=a+b\) of two positive integers \(a\) and \(b\) such that \(\Omega(a)\) and \(\Omega(b)\) have the same parity? The same question is asked with \(\Omega(a)\) and \(\Omega(b)\) having different parities. Questions of these type were raised in the author's previous paper [\textit{A. Gica}, A conjecture which implies the theorem of Gauss-Heegner-Stark-Baker, Bull. Math. Soc. Sci. Math. Roum., Nouv. Sér. 41, 111--118 (1998; Zbl 0968.11037)]. In Theorem 1, it is shown that every prime number \(p\equiv 7\pmod 8\) can be written in the form \(p=a+b^2\), with \(a\) and \(b\) positive integers and \(\Omega (b)\) even. The proof of this theorem is elementary and employs an argument similar to the standard one used to show that there are only finitely many rings \(\mathbb{Z}[\sqrt d]\) with \(d\equiv 2,3\pmod 4\) which are norm Euclidean (Exercise 2.4.8 in \textit{J. Esmonde} and \textit{M. Ram Murty}, Problems in Algebraic Number Theory, Springer-Verlag, New York (1999; Zbl 0911.11001)]. In Theorem 2, it is shown that if \(n>3\) then \(n\) can be written as \(n=a+b\) with \(a\) and \(b\) positive integers having \(\Omega(a)\equiv \Omega(b)\pmod 2\). The proof of this theorem is also elementary but very clever. First of all, using the fact that \(\Omega\) is totally additive, it suffices to show that the statement holds for \(n=2\), 9 or \(n=p>3\) a prime number. Secondly, since every prime number \(p\equiv 1 \pmod 4\) is a sum of two squares, it follows that it suffices to consider the case in which \(n=p\) is a prime which is congruent to 3 modulo 4. Assuming that the theorem does not hold for some prime number \(p>3\) with \(p\equiv 3\pmod 4\), the author proves that every prime number \(q<p\) must then be a non-quadratic residue modulo \(p\). This is achieved by induction (on the primes \(q<p)\) and by evaluating the relation \(\Omega(a) \not\equiv \Omega(b)\pmod 2\) in various representations of \(p\) in the form \(p=a+b\). The contradiction is achieved by using the fact that if \(p>5\) then there exists a prime \(q<p\) which is a quadratic residue modulo \(p\) (Lemma 1). Finally, in Theorem 3, the author shows that if \(n>2\) and \(n\neq 5.10\), then \(n\) can be written in the form \(n=a+b\) with \(a\) and \(b\) positive and \(\Omega(a)\not \equiv\Omega(b) \pmod 2\). The proof of this statement is very similar to the proof of Theorem 1. The author also points out some connections between the questions treated in this paper and quadratic fields with small class numbers.
    0 references
    0 references
    representation problems
    0 references
    number of prime divisors counted with multiplicities
    0 references
    parity
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references