Vol. 23, No. 11 (2000) 753–758 S0161171200003562
© Hindawi Publishing Corp.
ON CHARACTERIZATIONS OF A CENTER GALOIS EXTENSION
GEORGE SZETO and LIANYONG XUE (Received 16 June 1999)
Abstract.LetBbe a ring with 1,Cthe center ofB,Ga finite automorphism group ofB, andBGthe set of elements inBfixed under each element inG. Then, it is shown thatBis a center Galois extension ofBG(that is,Cis a Galois algebra overCGwith Galois group G|CG) if and only if the ideal ofBgenerated by{c−g(c)|c∈C}isBfor eachg≠1 in G. This generalizes the well known characterization of a commutative Galois extensionC thatCis a Galois extension ofCGwith Galois groupGif and only if the ideal generated by {c−g(c)|c∈C}isCfor eachg≠1 inG. Some more characterizations of a center Galois extensionBare also given.
Keywords and phrases. Galois extensions, center Galois extensions, central extensions, Ga- lois central extensions, Azumaya algebras, separable extensions,H-separable extensions.
2000 Mathematics Subject Classification. Primary 16S30, 16W20.
1. Introduction. LetCbe a commutative ring with 1,Ga finite automorphism group ofCandCGthe set of elements inCfixed under each element inG. It is well known that a commutative Galois extensionC is characterized in terms of the ideals generated by{c−g(c)|c∈C}forg≠1 inG, that isCis a Galois extension with Galois groupG if and only if the ideal generated by{c−g(c)|c∈C}isC for eachg≠1 inG(see [3, Proposition 1.2, page 80]). A natural generalization of a commutative Galois extension is the notion of a center Galois extension, that is, a noncommutative ringBwith a finite automorphism groupGand centerCis called a center Galois extension ofBG with Galois groupGifC is a Galois extension ofCG with Galois group G|C G. Ikehata (see [4, 5]) characterized a center Galois extension with a cyclic Galois group G of prime order in terms of a skew polynomial ring. Then, the present authors generalized the Ikehata characterization to center Galois extensions with Galois groupG of any cyclic order [7] and to center Galois extensions with any finite Galois groupG [8].
The purpose of the present paper is to generalize the above characterization of a commutative Galois extension to a center Galois extension. We shall show thatBis a center Galois extension ofBGif and only if the ideal ofBgenerated by{c−g(c)|c∈C}
isBfor each g≠1 in G. A center Galois extensionBis also equivalent to each of the following statements:
(i) Bis a Galois central extension ofBG, that is,B=BGCwhich isG-Galois extension ofBG.
(ii) Bis a Galois extension ofBGwith a Galois system{bi∈B, ci∈C, i=1,2,...,m}
for some integerm.
(iii) the ideal of the subringBGC generated by{c−g(c)|c∈C}isBGC for each g≠1 inG.
2. Definitions and notations. Throughout this paper,Bwill represent a ring with 1,G= {g1=1, g2,...,gn}an automorphism group ofBof ordernfor some integer n,Cthe center ofB, BGthe set of elements inBfixed under each element inG, and B∗Ga skew group ring in which the multiplication is given bygb=g(b)gforb∈ Bandg∈G.
Bis called aG-Galois extension ofBGif there exist elements{ai,bi∈B, i=1,2,...,m}
for some integermsuch thatm
i=1aig(bi)=δ1,g. Such a set{ai,bi}is called aG-Galois system forB.Bis called a center Galois extension ofBG ifCis a Galois algebra over CGwith Galois groupG|CG.Bis called a central extension ofBGifB=BGC, andB is called a Galois central extension ofBGifB=BGCis a Galois extension ofBGwith Galois groupG.
Let A be a subring of a ring B with the same identity 1. We denote VB(A) the commutator subring of Ain B. We callB a separable extension of Aif there exist {ai,bi∈B, i=1,2,...,mfor some integerm}such that
aibi=1,and
bai⊗bi= ai⊗bibfor allb∈Bwhere⊗is overA.Bis calledH-separable extension ofAif B⊗ABis isomorphic to a direct summand of a finite direct sum ofBas aB-bimodule.
Bis called centrally projective overAifBis a direct summand of a finite direct sum ofAas aA-bimodule.
3. The characterizations. In this section, we denoteJj(C)= {c−gj(c)|c∈C}. We shall show thatBis a center Galois extension ofBGif and only ifB=BJj(C), the ideal ofBgenerated byJj(C),for eachgj≠1 inG. Some more characterizations of a center Galois extensionBare also given. We begin with a lemma.
Lemma3.1. IfB=BJj(C)for eachgj≠1inG(that is,j≠1), then
(1)Bis a Galois extension ofBGwith Galois groupGand a Galois system{bi∈B;ci∈ C, i=1,2,...,m}for some integerm.
(2)Bis a centrally projective overBG. (3)B∗GisH-separable overB.
(4)VB∗G(B)=C.
Proof. (1) SinceB =BJj(C) for each j ≠1, there exist
b(j)i ∈B, ci(j) ∈C, i= 1,2,...,mj
for some integermj, j=2,3,...,nsuch thatmj i=1bi(j)
ci(j)−gj ci(j)
=1.
Therefore, mj
i=1b(j)i ci(j) =1+mj i=1b(j)i gj
c(j)i
. Letb(j)mj+1= −mj i=1b(j)i gj
ci(j) and cm(j)j+1=1. Thenmj+1
i=1 b(j)i ci(j)=1 andmj+1 i=1 b(j)i gj
ci(j)
=0. Letbi2,i3,...,in=b(2)i2 bi(3)3
···b(n)in andci2,i3,...,in=c(2)i2 ci(3)3 ...ci(n)n forij=1,2,...,mj+1 andj=2,3,...,n. Then
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
bi2,i3,...,inci2,i3,...,in=
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
bi(2)2 b(3)i3 ···bi(n)n c(2)i2 ci(3)3 ···c(n)in
=
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
b(2)i2 ci(2)2 b(3)i3 c(3)i3 ···b(n)in ci(n)n
=
m2+1 i2=1
b(2)i2 c(2)i2
m3+1 i3=1
b(3)i3 c(3)i3 ···
mn+1 in=1
bi(n)n ci(n)n =1 (3.1)
and for eachj≠1
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
bi2,i3,...,ingj(ci2,i3,...,in)
=
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
bi(2)2 b(3)i3 ···bi(n)n gj
c(2)i2 ci(3)3 ···ci(n)n
=
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
bi(2)2 b(3)i3 ···bi(n)n gj
c(2)i2 gj
ci(3)3
···gj
ci(n)n
=
m2+1 i2=1
m3+1 i3=1
···
mn+1 in=1
bi(2)2 gj
ci(2)2 bi(3)3 gj
ci(3)3
···b(n)in gj
ci(n)n
=
m2+1 i2=1
b(2)i2 gj
c(2)i2 m3+1
i3=1
b(3)i3 gj
c(3)i3
···
mn+1 in=1
b(n)in gj
ci(n)n
=0.
(3.2)
Thus,{bi2,i3,...,in∈B; ci2,i3,...,in∈C, ij=1,2,...,mj+1 andj=2,3,...,n}is a Galois system forB. This complete the proof of (1).
(2) By (1),B is a Galois extension ofBG with a Galois system {bi∈B, ci∈C, i= 1,2,...,m}for some integerm. Letfi:B→BGgiven byfi(b)=n
j=1gj(cib)for allb∈ B, i=1,2,...,m. Then it is easy to check thatfiis a homomorphism asBG-bimodule and b=m
i=1bicib =n
j=1m
i=1bigj(ci)gj(b)=m
i=1bin
j=1gj(cib)=m
i=1bifi(b) for allb∈B. Hence{bi;fi, i=1,2,...,m}is a dual bases forBasBG-bimodule, and so Bis finitely generated and projective asBG-bimodule. Therefore,Bis a direct summand of a finite direct sum ofBGas aBG-bimodule. ThusBis centrally projective overBG.
(3) By (1),Bis a Galois extension ofBGwith Galois groupG. HenceB∗GHomBG(B, B)[2, Theorem 1]. By (2),Bis centrally projective overBG. Thus,B∗G(HomBG(B,B)) isH-separable overB[6, Proposition 11].
(4) We first claim thatVB∗G(C)=B. Clearly,B⊂VB∗G(C). Letn
j=1bjgjin VB∗G(C) for somebj∈B. Thenc(n
j=1bjgj)=(n
j=1bjgj)cfor eachc∈ C, socbj=bjgj(c), that is,bj(c−gj(c))=0 for eachgj∈Gandc∈C. SinceB=BJj(C)for eachgj≠1, there existb(j)i ∈B andci(j)∈C, i=1,2,...,msuch thatm
i=1b(j)i
ci(j)−gj ci(j)
= 1. Hence bj=m
i=1b(j)i
ci(j)−gj ci(j)
bj=m
i=1bi(j)bj
ci(j)−gj ci(j)
=0 for each gj≠1. This implies thatn
j=1bjgj=b1∈B. HenceVB∗G(C)⊆B, and soVB∗G(C)=B.
Therefore,VB∗G(B)⊂VB∗G(C)=B. ThusVB∗G(B)=VB(B)=C.
We now show some characterizations of a center Galois extensionB.
Theorem3.2. The following statements are equivalent.
(1)Bis a center Galois extension ofBG. (2)B=BJj(C)for eachgj≠1inG.
(3)Bis a Galois extension ofBGwith a Galois system{bi∈B, ci∈C, i=1,2,...,m}
for some integerm.
(4)Bis a Galois central extension ofBG. (5)BGC=BGCJj(C)for eachgj≠1inG.
Proof. (1)⇒(2). By hypothesis,C is a Galois extension ofCG with Galois group G|C G. HenceC=CJj(C)for eachgj≠1 in G[3, Proposition 1.2, page 80]. Thus, B=BJj(C)for eachgj≠1 inG.
(2)⇒(1). SinceB=BJjCfor eachgj≠1 inG, B∗GisH-separable overBby Lemma 3.1(3) and VB∗G(B)=C by Lemma 3.1(4). Thus C is a Galois extension ofCG with Galois groupG|CGby [1, Proposition 4].
(1)⇒(3). This is Lemma 3.1(1).
(3)⇒(1). SinceBis Galois extension ofBGwith a Galois system{bi∈B, ci∈C, i= 1,2,...,m}for some integerm, we havem
i=1bigj(ci)=δ1,g. Hencem
i=1bi(ci−gj(ci))
=1 for eachgj≠1in G. So for everyb∈B,b=m
i=1bbi(ci−gj(ci))∈BJ(C)j . There- fore,B=BJi(C)for each gi≠1in G. Thus,B is a center Galois extension ofBG by (2)⇒(1).
(1)⇒(4). SinceCis a Galois algebra with Galois groupG|CG,BandBGCare Galois extensions ofBGwith Galois groupG|BGCG. Noting thatBGC⊂B, we haveB=BGC, that is,Bis a central extension ofBG. ButBis a Galois extension ofBG, soBis a Galois central extension ofBG.
(4)⇒(1). By hypothesis,B=BGC is a Galois extension ofBG. Hence there exists a Galois system{ai;bi∈B, i=1,2,...,m}for some integermsuch thatm
i=1aigj(bi)= δ1,j. ButB=BGC, soai=nai
k=1b(aki)c(aki)andbi=nbi
l=1bl(bi)cl(bi)for somea(aki), b(bl i) inBGandck(ai), c(bl i)in C, k=1,2...,nai, l=1,2,...,nbi, i=1,2,...,m. Therefore,
δ1,j= m i=1
aigj(bi)= m i=1
nai
k=1
b(aki)ck(ai)gj nbi
l=1
b(bl i)cl(bi)
= m i=1
nai
k=1
bk(ai)ck(ai)
nbi
l=1
bbligj
c(bl i)
= m i=1
nai
k=1 nbi
l=1
b(aki)ck(ai)b(bl i) gj
cl(bi) .
(3.3)
This shows that
bk,l(ai,bi)=b(aki)ck(ai)bl(bi)∈B; ck,l(ai,bi)=cl(bi)∈C, k=1,2,...,nai, l= 1,2,...,nbi, i=1,2,...,m
is a Galois system forB. Thus,Bis a center Galois extension ofBGby (3)⇒(1).
(1)⇒(5). SinceBis a center Galois extension ofBG,B=BJ(C)j for eachgj≠1 inG by (1)⇒(2) andB=BGCby (1)⇒(4). Thus, BGC=BGCJ(C)j for eachgj≠1 inG.
(5)⇒(1). SinceBGC=BGCJj(C)for eachgj≠1 inG,B=BJj(C)for eachgj≠1 inG.
Thus,Bis a center Galois extension ofBGby (2)⇒(1).
The characterization of a commutative Galois extensionC in terms of the ideals generated by {c−g(c)| c ∈C} for g ≠1 in G is an immediate consequence of Theorem 3.2.
Corollary3.3. A commutative ringC is a Galois extension ofCG if and only if C=CJj(C), the ideal generated by{c−gj(c)|c∈C}isCfor eachgj≠1inG.
Proof. LetB=Cin Theorem 3.2. Then, the corollary is an immediate consequence of Theorem 3.2(2).
By Theorem 3.2, we derive several characterizations of a Galois centeral extensionB.
Corollary3.4. IfBis a central extension ofBG(that is,B=BGC), then the following statements are equivalent.
(1)Bis a Galois extension ofBG. (2)Bis a center Galois extension ofBG. (3)B∗GisH-separable overB.
(4)B=CJj(B)for eachgj≠1inG.
(5)B=BJj(B)for eachgj≠1inG.
Proof. (1)⇐⇒(2). This is given by (1)⇐⇒(4) in Theorem 3.2.
(2)⇒(3). This is Lemma 3.1(3).
(3)⇒(1). SinceB∗GisH-separable overB,Bis a Galois extension ofBG[1, Propo- sition 2].
SinceB=BGCby hypothesis, it is easy to see thatJj(B)=BGJj(C)for eachgjinG. Thus, B=CJj(B), B=BJ(B)j , andB=BJj(C)are equivalent. This implies that (2)⇐⇒(4)⇐⇒(5) by Theorem 3.2(2).
We call a ringBthe DeMeyer-Kanzaki Galois extension ofBGifBis an AzumayaC- algebra andBis a center Galois extension ofBG(for more about the DeMeyer-Kanzaki Galois extensions, see [2]). Clearly, the class of center Galois extensions is broader than the class of the DeMeyer-Kanzaki Galois extensions. We conclude the present paper with two examples. (1) The DeMeyer-Kanzaki Galois extension ofBG and (2) a center Galois extension ofBG, but not the DeMeyer-Kanzaki Galois extension ofBG.
Example3.5. LetCbe the field of complex numbers, that is, C=R+R√
−1 where Ris the field of real numbers,B=C[i,j,k]the quaternion algebra overC, andG= {1,g|g(c1+cii+cjj+ckk)=g(c1)+g(ci)i+g(cj)j+g(ck)kfor eachb=c1+cii+
cjj+ckk∈C[i,j,k]andg(u+v√
−1)=u−v√
−1 for eachc=u+v√
−1∈C}. Then (1) The center ofBisC.
(2)Bis an AzumayaC-algebra.
(3)Cis a Galois extension ofCG with Galois groupG|CGand a Galois system {a1=1/√
2, a2=(1/√ 2)√
−1; b1=1/√
2, b2= −(1/√ 2)√
−1}.
(4)Bis the DeMeyer-Kanzaki Galois extension ofBGby (2) and (3).
(5)BG=R[i,j,k].
(6)B=BGC, soBis a centeral extension ofBG. (7)J(C)g =R√
−1.
(8)B=BJg(C)since 1= −√
−1√
−1∈BJg(C). (9)J(B)g =R√
−1+R√
−1i+R√
−1j+R√
−1k.
(10)B=CJg(B).
Example3.6. By replacing in Example 3.5 the field of complex numbers Cwith the ringC=Z⊕ZwhereZis the ring of integers,g(a,b)=(b,a)for all(a,b)∈C, andG= {1,g|g(c1+cii+cjj+ckk)=g(c1)+g(ci)i+g(cj)j+g(ck)kfor eachb= c1+cii+cjj+ckk∈B=C[i,j,k]}. Then
(1) The center ofBisC.
(2)C is a Galois extension ofCG with Galois groupG|C G and a Galois system {a1=(1,0), a2=(0,1); b1=(1,0), b2=(0,1)}.
(3)Bis not an AzumayaC-algebra (for 1/2∉C), and soBis not the DeMeyer-Kanzaki Galois extension ofBG.
(4)CG= {(a,a)|a∈Z} Z.
(5)BG=CG[i,j,k].
(6)B=BGC, soBis a central extension ofBG. (7)J(C)g = {(a,−a)|a∈Z} =Z(1,−1).
(8)B=BJg(C)since 1=(1,1)=(1,−1)(1,−1)∈BJG(C). (9)J(B)g =Z(1,−1)+Z(1,−1)i+Z(1,−1)j+Z(1,−1)k.
(10)B=CJg(B).
References
[1] R. Alfaro and G. Szeto,The centralizer onH-separable skew group rings, Rings, exten- sions, and cohomology (Evanston, IL, 1993) (New York), Dekker, 1994, pp. 1–7.
MR 95g:16027. Zbl 812.16038.
[2] F. R. DeMeyer,Some notes on the general Galois theory of rings, Osaka J. Math.2(1965), 117–127. MR 32#128. Zbl 143.05602.
[3] F. R. DeMeyer and E. Ingraham,Separable Algebras over Commutative Rings, Springer- Verlag, Berlin, 1971. MR 43#6199. Zbl 215.36602.
[4] S. Ikehata,OnH-separable polynomials of prime degree, Math. J. Okayama Univ.33(1991), 21–26. MR 93g:16043. Zbl 788.16022.
[5] S. Ikehata and G. Szeto,OnH-skew polynomial rings and Galois extensions, Rings, exten- sions, and cohomology (Evanston, IL, 1993) (New York), Dekker, 1994, pp. 113–121.
MR 95j:16033. Zbl 815.16009.
[6] K. Sugano,Note on separability of endomorphism rings, J. Fac. Sci. Hokkaido Univ. Ser. I 21(1970/71), 196–208. MR 45#3465. Zbl 236.16003.
[7] G. Szeto and L. Xue,On the Ikehata theorem forH-separable skew polynomial rings, Math.
J. Okayama Univ.40(1998), 27–32, 2000.
[8] ,The general Ikehata theorem forH-separable crossed products, Internat. J. Math.
Math. Sci., Vol 25, to appear, 1999.
Szeto: Department of Mathematics, Bradley University, Peoria, Illinois61625, USA E-mail address:[email protected]
Xue: Department of Mathematics, Bradley University, Peoria, Illinois61625, USA E-mail address:[email protected]
Mathematical Problems in Engineering
Special Issue on
Modeling Experimental Nonlinear Dynamics and Chaotic Scenarios
Call for Papers
Thinking about nonlinearity in engineering areas, up to the 70s, was focused on intentionally built nonlinear parts in order to improve the operational characteristics of a device or system. Keying, saturation, hysteretic phenomena, and dead zones were added to existing devices increasing their behavior diversity and precision. In this context, an intrinsic nonlinearity was treated just as a linear approximation, around equilibrium points.
Inspired on the rediscovering of the richness of nonlinear and chaotic phenomena, engineers started using analytical tools from “Qualitative Theory of Di
fferential Equations,”
allowing more precise analysis and synthesis, in order to produce new vital products and services. Bifurcation theory, dynamical systems and chaos started to be part of the mandatory set of tools for design engineers.
This proposed special edition of the Mathematical Prob-
lems in Engineering aims to provide a picture of the impor-tance of the bifurcation theory, relating it with nonlinear and chaotic dynamics for natural and engineered systems.
Ideas of how this dynamics can be captured through precisely tailored real and numerical experiments and understanding by the combination of specific tools that associate dynamical system theory and geometric tools in a very clever, sophis- ticated, and at the same time simple and unique analytical environment are the subject of this issue, allowing new methods to design high-precision devices and equipment.
Authors should follow the Mathematical Problems in Engineering manuscript format described at
http://www .hindawi.com/journals/mpe/. Prospective authors shouldsubmit an electronic copy of their complete manuscript through the journal Manuscript Tracking System at
http://mts.hindawi.com/
according to the following timetable:
Manuscript Due December 1, 2008 First Round of Reviews March 1, 2009 Publication Date June 1, 2009
Guest Editors
José Roberto Castilho Piqueira,
Telecommunication and Control Engineering Department, Polytechnic School, The University of São Paulo, 05508-970 São Paulo, Brazil;
[email protected]
Elbert E. Neher Macau,
Laboratório Associado de Matemática Aplicada e Computação (LAC), Instituto Nacional de Pesquisas Espaciais (INPE), São Josè dos Campos, 12227-010 São Paulo, Brazil ; [email protected]
Celso Grebogi,Center for Applied Dynamics Research, King’s College, University of Aberdeen, Aberdeen AB24 3UE, UK; [email protected]
Hindawi Publishing Corporation http://www.hindawi.com