Sunday, 21 July 2019

finite fields - Roots of polynomials of the form xqxalpha over mathbbFqm



Let q be a prime power and Fq a finite field with q elements.
Moreover let Fqm be an extension field. For every αFqm define
the polynomial

gα(x)=xqxαFqm[x].
My questions:




  1. For which α (and m) the polynomial
    gα(x) has at least one root in Fqm?


  2. Which is the splitting field of gα?




Here my thoughts.





  1. If β is a root of gα, then the set of roots of gα is
    β+Fq:={β+λ|λFq}.
    This means that if gα has one root in Fqm then all the roots belongs to Fqm.


  2. If α0 then βFq.




Any help or reference will be appreciated.
Thanks.




Alessandro



EDIT



It seems by computation with Macaulay2, for every αFqm there are only 2 possibilities for the polynomial gα.




  1. gα splits into linear factors.


  2. There exist λFq, γ1,,γqmpFqm (where p is the characteritic of the field), s.t.

    gα(x)=qmpi=1(xpλxγi).
    But I don't know how to prove it.



Answer



Partial answer and/or something that is too long to fit into a comment.



Consider the relative trace function
trm1:FqmFq,xx+xq+xq2++xqm1.

We easily see (or know) that trm1 is Fq-linear, and surjective. Furthermore,
trm1(zq)=trm1(z)
for all zFqm. Therefore, again for all zFqm, we have trm1(zqz)=0. The mapping
P:FqmFqm,xxqx
is also Fq-linear. Clearly ker(P)=Fq, so rank-nullity theorem implies that im(P) has dimension m1. But we just saw that im(P)ker(trm1). Therefore we get equality



im(P)=ker(trm1).
Together with your observation this implies that





gα(x) has a zero in the field Fqm if and only if trm1(α)=0.




As you observed, when this holds all the q zeros of gα(x) will be in the field Fqm.






The factorization you observed can be described as follows.

Assume that trm1(α)0. Let z be a zero of gα(x) in some extension field of Fqm. Let's consider the Galois conjugates of z over Fqm. These are gotten by iterating the Frobenius automorphism F:zzqm. But we know that F(α)=α and that
F(z)=zqm=zqmgα(z)=z+α.
An easy induction then proves that for all i=1,2,,p we have
Fi(z)=z+iα.
In particular the lowest power of F that maps z to itself is Fp. Therefore the minimal polynomial of z over Fqm is
mz(x):=p1i=0(xziα).
Let's denote T(x)=xpx=p1i=0(xi)Fp[x]. We see that
T(xzα)=p1i=0(xzαi)=αpmz(x).
But T is additive, i.e. T(x+y)=T(x)+T(y) for all x,y¯Fq, so
mz(x)=αpT(xzα)=αpT(x/α)αpT(z/α)=xpαp1xαpT(z/α).
In particular T(z/α) must be an element of Fqm. These are the factors that you see in your Macaulay output. In particular your λ=αp1.


No comments:

Post a Comment

real analysis - How to find limhrightarrow0fracsin(ha)h

How to find lim without lhopital rule? I know when I use lhopital I easy get $$ \lim_{h\rightarrow 0}...