3
$\begingroup$

Here $\mathbb{F}_{q}$ means a finite field with $q = p^m$ elements where $p$ is the characteristic of the field in question and $\overline{\mathbb{F}_{q}}$ means its algebraic closure.

I am studying an article about the existence of a normal basis with a primitive element for finite extensions of $\mathbb{F}_{q}$. And there is this new definition for me, that I haven't found anywhere:

Let $\alpha \in \overline{\mathbb{F}_{q}}$, and let $\sigma$ the Frobenius Authomorphism ($\sigma(\alpha) = \alpha^q$). One can turn $\overline{\mathbb{F}_{q}}$ into a $\mathbb{F}_{q}[X]$-module as follows:

If $f \in \mathbb{F}_{q}[X]$, write $f = \sum_{i=0}^{k} a_{i}X^{i}$ and then define:

$f \circ \alpha := \sum_{i=0}^{k} a_{i}\sigma^{i}(\alpha)$

Which is essentially $f(\sigma)(\alpha)$.

We then have that

$\alpha \in \mathbb{F}_{q^n} \Leftrightarrow \alpha^{q^n} = \alpha \Leftrightarrow \sigma^{n}(\alpha) = \alpha \Leftrightarrow (X^n-1)\circ\alpha =0$,

and then the Annihilator of $\alpha$ over $\mathbb{F}_{q}[X]$ is non-trivial for every $\alpha$. Then one can define the, let's call here additive order ($Ord(\alpha)$) of $\alpha$ by the monic polynomial generating the Annihilator of $\alpha$. It's easy to see that $\alpha \in \mathbb{F}_{q^n} \Leftrightarrow Ord(\alpha)|X^n-1$.

The point is, in the article, it is said that:

If $\alpha \in \mathbb{F}_{q^n}$ has $Ord(\alpha) = g$, then $\alpha = h\circ\beta$, for some $\beta \in \mathbb{F}_{q^n}$ and $h = \dfrac{X^n-1}{g}$.

The point is, I just cannot prove this (it seems trivial since every source I look into just states this and does not prove).

Well, if the characteristic of $\mathbb{F}_{q^n}$ does not divide $n$, then it is clear for me, since $X^n-1$ is separable over $\mathbb{F}_{q}[X]$ and then $(g,h) = 1$, therefore there exists $s,t \in \mathbb{F}_{q}[X]$ such that

$sg + th = 1$, which implies $(sg + th)\circ \alpha = 1\circ \alpha$ and $ (sg)\circ \alpha + (th)\circ\alpha = (th)\circ\alpha = h\circ(t\circ\alpha) = \alpha$, defining $\beta = t \circ \alpha$ we get the result. $(sg)\circ \alpha = 0$ because $g$ is the order of $\alpha$ therefore $g\circ\alpha = 0$, and I used the fact that we have a module structure over $\mathbb{F}_{q}[X]$.

Now, if $n$ divide the characteristic of $\mathbb{F}_{q^n}$, I cannot prove by this way. Well, I think there is a simpler argument such that one doesn't need to split the proof between two cases. Can anyone help me with the second case or can anyone give me a hint or a new argument that helps me to solve this problem?

$\endgroup$
1

1 Answer 1

4
$\begingroup$

By the normal basis theorem, there exists $a \in \mathbb F_{q^n}$ such that $\{\sigma(a)\mid \sigma \in \operatorname{Gal}(\mathbb F_{q^n}/\mathbb F_q)\}$ spans $\mathbb F_{q^n}$ as $\mathbb F_q$-vector space. Since $\operatorname{Gal}(\mathbb F_{q^n}/\mathbb F_q)$ is generated by Frobenius, this means that $\mathbb F_{q^n}$ is spanned by the Frobenius powers of $a$, i.e. is generated as $\mathbb F_q[X]$-module by $a$.

Thus $\mathbb F_{q^n} \cong \mathbb F_q[X]/I$, where $I = (\operatorname{Ann}(a))$. Since $\mathbb F_{q^n}$ is killed by $X^n - 1$, we get $(X^n - 1) \subseteq I$. But $$|\mathbb F_q[X]/(X^n-1)| = q^n = |\mathbb F_{q^n}| = |\mathbb F_q[X]/I|,$$ so the inclusion $(X^n - 1) \subseteq I$ has to be an equality. That is, $$\mathbb F_{q^n} \cong \mathbb F_q[X]/(X^n-1),$$ as $\mathbb F_q[X]$-modules. The result now follows from the following lemma.

Lemma. Let $R$ be a PID, let $M = R/(f)$ for some $f \in R\setminus\{0\}$, and let $m \in M$. If $\operatorname{Ann}(m) = (g)$, then there exists $h \in R$ such that $gh = f$, and there exists $n \in M$ such that $hn = m$.

Proof. The first statement follows from the fact that $(f) \subseteq (g)$ since $\operatorname{Ann}(M) \subseteq \operatorname{Ann}(m)$. For the second statement, note that the sequence $$0 \to (h)/(f) \to R/(f) \to R/(h) \to 0$$ is exact. But multiplication by $g$ induces an isomorphism $R/(h) \to (g)/(f)$. Hence, the sequence above turns into $$0 \to (h)/(f) \to R/(f) \stackrel{\cdot g}\longrightarrow (g)/(f) \to 0.$$ This proves the second statement: if $gm = 0$, then $m$ is divisible by $h$. $\square$

$\endgroup$
1
  • 1
    $\begingroup$ Thanks, this is exactly the lemma I was looking for. Actually this works for me, since the article I'm studying is about the existence of a normal basis that the element $a$ such that $\lbrace \sigma(a) \rbrace$ spans $\mathbb{F}_{q^n}$ as a $\mathbb{F}_{q}$-vector space is also a primitive element of $\mathbb{F}_{q^n}$, so, using the normal basis theorem is allowed. Thank you, anyway. $\endgroup$
    – Lucas
    May 3, 2017 at 1:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.