Prime $p>2$ and a polynomial $Q$ with integer coefficients are such that there are no integers $1 \le i < j \le p-1$ for which $(Q(j)-Q(i))(jQ(j)-iQ(i))$ is divisible by $p$. What is the smallest possible degree of $Q$? (Proposed by Anton Trygub)
Problem
Source: Kyiv City MO 2022 Round 2, Problem 10.4
Tags: number theory, polynomial, algebra
laikhanhhoang_3011
31.01.2022 08:14
MS_Kekas wrote: Prime $p>2$ and a polynomial $Q$ with integer coefficients are such that there are no integers $1 \le i < j \le p-1$ for which $(Q(j)-Q(i))(jQ(j)-iQ(i))$ is divisible by $p$. What is the smallest possible degree of $Q$? (Proposed by Anton Trygub)
see here
Assume $Q$ is the polynomial with the smallest degree satisfied the statement
Claim 1: $deg(Q) = p-2$
Proof:
$\bullet$ $deq(Q)>p-1$. Then use the identity $x^p \equiv x \textrm{(mod p)}$, we get a contradiction with $deg(Q)$ is smallest
$\bullet$ $deq(Q)=p-1$. Then we can use the identity $x^{p-1} \equiv 1 \textrm{(mod p)}$ because the problem is just care about $\overline{1,p-1}$
$\bullet$ $deg(Q)<p-2$. We have two case to kill:
Case 1: $(Q(i),p)=1, \forall i=\overline{1,p-1}$
We have $\left\{Q(1),Q(2),..,Q(p-1) \right\}$ and $\left\{ 1\times Q(1),2\times Q(2),...,(p-1)\times Q(p-1)\right\}$ are two full residues mod p
So by Wilson theorem, we get: $\left\{\begin{matrix}Q(1)\times Q(2)\times ...\times Q(p-1)\equiv -1 (\textrm{mod p})
& \\
& \\ (1.Q(1))\times (2.Q(2))... \times ((p-1).Q(p-1))\equiv - 1 (\textrm{mod p})
\end{matrix}\right.$
But $$(1.Q(1))\times (2.Q(2))...\times ((p-1).Q(p-1)) \equiv (1.2...(p-1))\times \left [ Q(1).Q(2)...Q(p-1) \right ]\equiv (-1).(-1) \equiv 1 (\textrm{mod p})$$a contradiction
Case 2: Exist $i: Q(i) \vdots p$
Because $deg(Q)<p-2$, $deg(xQ(x)) \leq p-2$ .
So we have the identity: $\sum_{i=0}^{p-1}iQ(i)\equiv 0 (\textrm{mod p}) \leftrightarrow \sum_{i=1}^{p-1}iQ(i)\equiv 0 (\textrm{mod p}), (1)$
But exist $i:Q(i) \vdots p$ and $\left\{ 1\times Q(1),2\times Q(2),...,(i-1)Q(i-1),(i+1)Q(i+1),...(p-1)\times Q(p-1)\right\} \sim (1,2,...x-1,x+1,...,p-1)$
From $(1)$, we get: $1+2+...+(p-1)-x \equiv 0 (\textrm{mod p}) \leftrightarrow x\equiv 0 (\textrm{mod p}) $, a contradiction
Q.E.D
I haven't got any ideas, can someone help me ?
one years back, is it just $Q(x)=x^{p-2}+1$ ?
P/s: My 409-th post. This is a beautiful number for me