free resolution of $\mathbb{Z}$ as an $\mathbb{Z}[x]/(x^n-1)$-module

381 Views Asked by At

As the tittle says, I am having a bad time thinking on the construction of a free resolution of $\mathbb{Z}$ as an $(\mathbb{Z}[x]/(x^n-1))$-module.

I know that I should give an exact sequence of the form (if $R= \mathbb{Z}[x]/(x^n-1)$):

$... \longrightarrow R^{I_n} \longrightarrow ...\longrightarrow R^{I_1} \longrightarrow R^{I_0} \longrightarrow \mathbb{Z} \longrightarrow 0$

being

$d_i : R^{I_i} \longrightarrow R^{I_{i-1}}$ for $i \ge 1 $

$d_0: R^{I_0} \longrightarrow \mathbb{Z}$

$e: \mathbb{Z} \longrightarrow 0$.

and $R^{I_i}$ being free $R$-modules.

Moreover they should satisfy the exact sequence condition: $Im(d_i)=Ker(d_{i-1})$.

I began thinking on the first application $d_0: R^{I_0} \longrightarrow \mathbb{Z}$. But I had some troubles finding it. I thought sending

$a_0 + a_1 \bar{x} + a_2 \bar{x}^2+...+a_{n-1} \bar{x}^{n-1} \mapsto a_0$

because I need $Im(d_0)=Ker(e) = \mathbb{Z}$. But it gets harder to find $d_1$.

Then I thought:

$a_0 + a_1 \bar{x} + a_2 \bar{x}^2+...+a_{n-1} \bar{x}^{n-1} \mapsto \sum a_i$

but again it gets hard.

Any hint/help?

EDIT: I chose $R^{I_0}$ to be $R$.

1

There are 1 best solutions below

10
On BEST ANSWER

Consider the following sequence of $R$-modules:

$$ \ldots \longrightarrow R \longrightarrow R \longrightarrow R \longrightarrow \mathbb{Z} \rightarrow 0 . $$ We denote by $d_i$ the usual map. We define $d_0: R \rightarrow \mathbb{Z}: f(\overline{x}) \rightarrow f(1).$ Then, the kernel of $d_0$ corresponds to the polynomials in $R$ such that $f(1) = 0$. Consider now $d_1: R \rightarrow R: g(\overline{x})\rightarrow (\overline{x-1})g(\overline{x}). $

Notice that, per construction, the image of $d_1$ is the kernel of $d_0$. The kernel of $d_1$ is the set of functions such that $g(\overline{x})(\overline{x-1})=0$. Try to complete the sequence, using the fact that $(1+x+\ldots + x^{n-1})(x-1) = x^n-1.$