Cauchy Davenport Theorem
Recently when I was solving some random math olympiad problems, I found this interesting theorem and decide to share it.
Given two sets \(A,B\) , define
If the two sets are defined over real numbers, we can prove that
It is not difficult to show that this is true, as we are able to order the elements in \(A\) and \(B\) as \(a_1<a_2<\cdots<a_n,b_1<b_2<\cdots<b_m\) . Then
Thus, there are at least \(n+m-1\) elements in \(A+B\) ! The equality holds only when \(A,B\) are arithmetic sequences with the same common difference. The original theorem states the inequality when \(A,B\) are in a cyclic group with prime order \(\mathbb{Z}/{p\mathbb{Z}}\) .
Lemma 1: If \(G\) is a finite abelian group and \(A,B\) are nonempty subsets of \(G\) such that \(|A|+|B|> |G|\), then \(A+B=G\) .
Proof: This is equivalent to proving that for any \(g\in G\) , there exist \(a\in A,b\in B\) such that \(a+b=g\). Define \(g-B=\{g-b|b\in B\}\) , then \(|g-B|=|B|\) . We now have \(|A|+|g-B|>|G|\) . Thus \(A\cap g-B\ne\emptyset\) , pick \(a\in A\cap g-B\) , we get a solution for \(a+b=g\) .