I am trying to understand the proof of the following question:
If automorphisms of $G$ are cyclic, then $G$ is abelian.
I found the following answer here If $G$ is non-Abelian, show that $Aut(G)$ is not cyclic. :
Suppose that ${\rm Aut}\,(G)$ is cyclic and let $f$ its generator, let $x,y\in G$ such that $y$ does not commute with $yxy^{-1}$, (since $G$ is not commutative, you can find $z$ such that $z$ does not commute with $y$, write $x=y^{-1}zy, yxy^{-1}=z$) there exists $n$ such that $c_x(z)=xzx^{-1}=f^n, m: c_y(z)=yzy^{-1}=f^m$, this implies that $c_x$ commutes with $c_y$, we deduce that $c_y(c_x(x))=c_x(x))$, i.e $yxy^{-1}=xyxy^{-1}x^{-1}$, i.e $(yxy^{-1})x=x(yxy^{-1})$ i.e $x$ commutes with $yxy^{-1}$. Contradiction.
But I really do not understand this solution, could someone clarify it for me please?
EDIT: I am not searching for a proof only, I am searching for an easy proof that do not require previous theorems and I think the above proof do this, so the suggested duplicate does not answer my question, I hope the question is reopened for answers please.