3
$\begingroup$

I am interested in whether there are any conclusion/conterexample that the sum&difference of finite integer sets by more times is smaller than the sum&difference by less times.

The question I am actually thinking about is the following: Is there any example of finite integer set $U,V$, s.t. $U-V=\{1,2,...,m\}$, $U+U+V+V\neq\{1,2,...,m\}\pmod m$ for some m.

I think I knew nothing when comparing 4 sums and 2 difference, even the $U=V$ case or without modulo $m$ case.

when $U=\{x^2,x^2+1:x\in Z\}$, $V=\{x^2:x\in Z\}$, then $U-V=Z$, and $7*4^n \notin U+V+V$, this may show that 3 sums may be smaller than 2 difference? (though these are two infinite sets)

$\endgroup$
3
  • $\begingroup$ We have $2U+2V = 2(U+V)$ and $|U-V| \leq \frac{|U+V|^3}{|U||V|}$. Maybe we can find from this bounds on the doubling constant of $U+V$? Not sure if they are sufficient for anything, though $\endgroup$ Commented May 5 at 10:44
  • 5
    $\begingroup$ According to More differences than multiple sums, Imre Z. Ruzsa this is possible, even with $U=V$. $\endgroup$ Commented May 5 at 10:50
  • $\begingroup$ Thanks a lot! This is the result I'm exactly looking for. $\endgroup$
    – zzy
    Commented May 6 at 16:26

0

You must log in to answer this question.

Browse other questions tagged .