<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
<head>
<title>GAP (Wedderga) - Chapter 7: The basic theory behind Wedderga</title>
<meta http-equiv="content-type" content="text/html; charset=UTF-8" />
<meta name="generator" content="GAPDoc2HTML" />
<link rel="stylesheet" type="text/css" href="manual.css" />
</head>
<body>
<div class="chlinktop"><span class="chlink1">Goto Chapter: </span><a href="chap0.html">Top</a> <a href="chap1.html">1</a> <a href="chap2.html">2</a> <a href="chap3.html">3</a> <a href="chap4.html">4</a> <a href="chap5.html">5</a> <a href="chap6.html">6</a> <a href="chap7.html">7</a> <a href="chapBib.html">Bib</a> <a href="chapInd.html">Ind</a> </div>
<div class="chlinkprevnexttop"> <a href="chap0.html">Top of Book</a> <a href="chap6.html">Previous Chapter</a> <a href="chapBib.html">Next Chapter</a> </div>
<p><a id="X840E625A81FDAEC6" name="X840E625A81FDAEC6"></a></p>
<div class="ChapSects"><a href="chap7.html#X840E625A81FDAEC6">7 <span class="Heading">The basic theory behind <strong class="pkg">Wedderga</strong></span></a>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X815ECCD97B18314B">7.1 <span class="Heading">Group rings and group algebras</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X7FDD93FB79ADCC91">7.2 <span class="Heading">Semisimple group algebras</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X87273420791F220E">7.3 <span class="Heading">Wedderburn decomposition</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X87B6505C7C2EE054">7.4 <span class="Heading">Characters and primitive central idempotents</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X7A24D5407F72C633">7.5 <span class="Heading">Central simple algebras and Brauer equivalence</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X7FB21779832CE1CB">7.6 <span class="Heading">Crossed Products</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X828C42CD86AF605F">7.7 <span class="Heading">Cyclic Crossed Products</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X7869E2A48784C232">7.8 <span class="Heading">Abelian Crossed Products</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X80BABE5078A29793">7.9 <span class="Heading">Classical crossed products</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X84C98BB8859BBEE2">7.10 <span class="Heading">Cyclic Algebras</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X8099A8C784255672">7.11 <span class="Heading">Cyclotomic algebras</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X84A142407B7565E0">7.12 <span class="Heading">Numerical description of cyclotomic algebras</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X8310E96086509397">7.13 <span class="Heading">Idempotents given by subgroups</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X80C058BE81824B23">7.14 <span class="Heading">Shoda pairs</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X81DAF5267D30C83A">7.15 <span class="Heading">Strong Shoda pairs</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X84C694978557EFE5">7.16 <span class="Heading">Strongly monomial characters and strongly monomial groups</span></a>
</div>
<div class="ContSect"><span class="nocss"> </span><a href="chap7.html#X800D8C5087D79DC8">7.17 <span class="Heading">Cyclotomic Classes and Strong Shoda Pairs</span></a>
</div>
</div>
<h3>7 <span class="Heading">The basic theory behind <strong class="pkg">Wedderga</strong></span></h3>
<p>In this chapter we describe the theory that is behind the algorithms used by <strong class="pkg">Wedderga</strong>.</p>
<p>All the rings considered in this chapter are associative and have an identity.</p>
<p>We use the following notation: ℚ denotes the field of rationals and F_q the finite field of order q. For every positive integer k, we denote a complex k-th primitive root of unity by xi_k and so ℚ(xi_k) is the k-th cyclotomic extension of ℚ.</p>
<p><a id="X815ECCD97B18314B" name="X815ECCD97B18314B"></a></p>
<h4>7.1 <span class="Heading">Group rings and group algebras</span></h4>
<p>Given a group G and a ring R, the <em>group ring</em> RG over the group G with coefficients in R is the ring whose underlying additive group is a right R-module with basis G such that the product is defined by the following rule</p>
<p class="pcenter">
(gr)(hs)=(gh)(rs)
</p>
<p>for r,s in R and g, h in G, and extended to RG by linearity.</p>
<p>A <em>group algebra</em> is a group ring in which the coefficient ring is a field.</p>
<p><a id="X7FDD93FB79ADCC91" name="X7FDD93FB79ADCC91"></a></p>
<h4>7.2 <span class="Heading">Semisimple group algebras</span></h4>
<p>We say that a ring R is semisimple if it is a direct sum of simple left (alternatively right) ideals or equivalently if R is isomorphic to a direct product of simple algebras each one isomorphic to a matrix ring over a division ring.</p>
<p>By Maschke's Theorem, if G is a finite group then the group algebra FG is semisimple if and only the characteristic of the coefficient field F does not divide the order of G.</p>
<p>In fact, an arbitrary group ring RG is semisimple if and only if the coefficient ring R is semisimple, the group G is finite and the order of G is invertible in R.</p>
<p>Some authors use the notion semisimple ring for rings with zero Jacobson radical. To avoid confusion we usually refer to semisimple rings as semisimple artinian rings.</p>
<p><a id="X87273420791F220E" name="X87273420791F220E"></a></p>
<h4>7.3 <span class="Heading">Wedderburn decomposition</span></h4>
<p>If R is a <em>semisimple ring</em> (<a href="chap7.html#X7FDD93FB79ADCC91"><b>7.2</b></a>) then the <em>Wedderburn decomposition</em> of R is the decomposition of R as a direct product of simple algebras. The factors of this Wedderburn decomposition are called <em>Wedderburn components</em> of R. Each Wedderburn component of R is of the form Re for e a <em>primitive central idempotent</em> (<a href="chap7.html#X87B6505C7C2EE054"><b>7.4</b></a>) of R.</p>
<p>Let FG be a <em>semisimple group algebra</em> (<a href="chap7.html#X7FDD93FB79ADCC91"><b>7.2</b></a>). If F has positive characteristic, then the Wedderburn components of FG are matrix algebras over finite extensions of F. If F has zero characteristic then by the <em>Brauer-Witt Theorem</em> <a href="chapBib.html#biBY">[Yam74]</a>, the <em>Wedderburn components</em> of FG are <em>Brauer equivalent</em> (<a href="chap7.html#X7A24D5407F72C633"><b>7.5</b></a>) to <em>cyclotomic algebras</em> (<a href="chap7.html#X8099A8C784255672"><b>7.11</b></a>).</p>
<p>The main functions of <strong class="pkg">Wedderga</strong> compute the Wedderburn components of a semisimple group algebra FG, such that the coefficient field is either an abelian number field (i.e. a subfield of a finite cyclotomic extension of the rationals) or a finite field. In the finite case, the Wedderburn components are matrix algebras over finite fields and so can be described by the size of the matrices and the size of the finite field.</p>
<p>In the zero characteristic case each Wedderburn component A is <em>Brauer equivalent</em> (<a href="chap7.html#X7A24D5407F72C633"><b>7.5</b></a>) to a <em>cyclotomic algebra</em> (<a href="chap7.html#X8099A8C784255672"><b>7.11</b></a>) and therefore A is a (possibly fractional) matrix algebra over <em>cyclotomic algebra</em> and can be described numerically in one of the following three forms:</p>
<p class="pcenter">
[n,K],
</p>
<p class="pcenter">
[n,K,k,[d,\alpha,\beta]],
</p>
<p class="pcenter">
[n,K,k,[d_i,\alpha_i,\beta_i]_{i=1}^m, [\gamma_{i,j}]_{1\le i < j \le n} ],
</p>
<p>where n is the matrix size, K is the centre of A (a finite field extension of F) and the remaining data are integers whose interpretation is explained in <a href="chap7.html#X84A142407B7565E0"><b>7.12</b></a>.</p>
<p>In some cases (for the zero characteristic coefficient field) the size n of the matrix algebras is not a positive integer but a positive rational number. This is a consequence of the fact that the <em>Brauer-Witt Theorem</em> <a href="chapBib.html#biBY">[Yam74]</a> only ensures that each <em>Wedderburn component</em> (<a href="chap7.html#X87273420791F220E"><b>7.3</b></a>) of a semisimple group algebra is Brauer equivalent (<a href="chap7.html#X7A24D5407F72C633"><b>7.5</b></a>) to a <em>cyclotomic algebra</em> (<a href="chap7.html#X8099A8C784255672"><b>7.11</b></a>), but not necessarily isomorphic to a full matrix algebra of a cyclotomic algebra. For example, a Wedderburn component D of a group algebra can be a division algebra but not a cyclotomic algebra. In this case M_n(D) is a cyclotomic algebra C for some n and therefore D can be described as M_1/n(C) (see last Example in <code class="func">WedderburnDecomposition</code> (<a href="chap2.html#X7F1779ED8777F3E7"><b>2.1-1</b></a>)).</p>
<p>The main algorithm of <strong class="pkg">Wedderga</strong> is based on a computational oriented proof of the Brauer-Witt Theorem due to Olteanu <a href="chapBib.html#biBO">[Olt07]</a> which uses previous work by Olivieri, del Río and Simón <a href="chapBib.html#biBORS">[ORS04]</a> for rational group algebras of <em>strongly monomial groups</em> (<a href="chap7.html#X84C694978557EFE5"><b>7.16</b></a>).</p>
<p><a id="X87B6505C7C2EE054" name="X87B6505C7C2EE054"></a></p>
<h4>7.4 <span class="Heading">Characters and primitive central idempotents</span></h4>
<p>A <em>primitive central idempotent</em> of a ring R is a non-zero central idempotent e which cannot be written as the sum of two non-zero central idempotents of Re, or equivalently, such that Re is indecomposable as a direct product of two non-trivial two-sided ideals.</p>
<p>The <em>Wedderburn components</em> (<a href="chap7.html#X87273420791F220E"><b>7.3</b></a>) of a semisimple ring R are the rings of the form Re for e running over the set of primitive central idempotents of R.</p>
<p>Let FG be a <em>semisimple group algebra</em> (<a href="chap7.html#X7FDD93FB79ADCC91"><b>7.2</b></a>) and chi an irreducible character of G (in an algebraic closure of F). Then there is a unique Wedderburn component A=A_F(chi) of FG such that chi(A)ne 0. Let e_F(chi) denote the unique primitive central idempotent of FG in A_F(chi), that is the identity of A_F(chi), i.e.</p>
<p class="pcenter">
A_F(\chi)=FGe_F(\chi).
</p>
<p>The centre of A_F(chi) is F(chi)=F(chi(g):g in G), the <em>field of character values</em> of chi over F.</p>
<p>The map chi -> A_F(chi) defines a surjective map from the set of irreducible characters of G (in an algebraic closure of F) onto the set of Wedderburn components of FG.</p>
<p>Equivalently, the map chi -> e_F(chi) defines a surjective map from the set of irreducible characters of G (in an algebraic closure of F) onto the set of primitive central idempontents of FG.</p>
<p>If the irreducible character chi of G takes values in F then</p>
<p class="pcenter">
e_F(\chi) = e(\chi) = \frac{\chi(1)}{|G|} \sum_{g\in G} \chi(g^{-1}) g.
</p>
<p>In general one has</p>
<p class="pcenter">
e_F(\chi) = \sum_{\sigma \in Gal(F(\chi)/F)} e(\sigma \circ \chi).
</p>
<p><a id="X7A24D5407F72C633" name="X7A24D5407F72C633"></a></p>
<h4>7.5 <span class="Heading">Central simple algebras and Brauer equivalence</span></h4>
<p>Let K be a field. A <em>central simple K-algebra</em> is a finite dimensional K-algebra with center K which has no non-trivial proper ideals. Every central simple K-algebra is isomorphic to a matrix algebra M_n(D) where D is a division algebra (which is finite-dimensional over K and has centre K). The division algebra D is unique up to K-isomorphisms.</p>
<p>Two central simple K-algebras A and B are said to be <em>Brauer equivalent</em>, or simply <em>equivalent</em>, if there is a division algebra D and two positive integers m and n such that A is isomorphic to M_m(D) and B is isomorphic to M_n(D).</p>
<p><a id="X7FB21779832CE1CB" name="X7FB21779832CE1CB"></a></p>
<h4>7.6 <span class="Heading">Crossed Products</span></h4>
<p>Let R be a ring and G a group.</p>
<p><strong class="button">Intrinsic definition</strong>. A <em>crossed product</em> <a href="chapBib.html#biBP">[Pas89]</a> of G over R (or with coefficients in R) is a ring R*G with a decomposition into a direct sum of additive subgroups</p>
<p class="pcenter">
R*G = \bigoplus_{g \in G} A_g
</p>
<p>such that for each g,h in G one has:</p>
<p>* A_1=R (here 1 denotes the identity of G),</p>
<p>* A_g A_h = A_gh and</p>
<p>* A_g has a unit of R*G.</p>
<p><strong class="button">Extrinsic definition</strong>. Let Aut(R) denote the group of automorphisms of R and let R^* denote the group of units of R.</p>
<p>Let a:G -> Aut(R) and t:G x G -> R^* be mappings satisfying the following conditions for every g, h and k in G:</p>
<p>(1) a(gh)^-1 a(g) a(h) is the inner automorphism of R induced by t(g,h) (i.e. the automorphism x-> t(g,h)^-1 x t(g,h)) and</p>
<p>(2) t(gh,k) t(g,h)^k = t(g,hk) t(h,k), where for g in G and x in R we denote a(g)(x) by x^g.</p>
<p>The <em>crossed product</em> <a href="chapBib.html#biBP">[Pas89]</a> of G over R (or with coefficients in R), action a and twisting t is the ring</p>
<p class="pcenter">
R*_a^t G = \bigoplus_{g\in G} u_g R
</p>
<p>where u_g : gin G is a set of symbols in one-to-one correspondence with G, with addition and multiplication defined by</p>
<p class="pcenter">
(u_g r) + (u_g s) = u_g(r+s), \quad (u_g r)(u_h s) = u_{gh} t(g,h) r^h s
</p>
<p>for g,h in G and r,sin R, and extended to R*_a^t G by linearity.</p>
<p>The associativity of the product defined is a consequence of conditions (1) and (2) <a href="chapBib.html#biBP">[Pas89]</a>.</p>
<p><strong class="button">Equivalence of the two definitions</strong>. Obviously the crossed product of G over R defined using the extrinsic definition is a crossed product of G over u_1 R in the sense of the first definition. Moreover, there is r_0 in R^* such that u_1r_0 is the identity of R*_a^t G and the map r -> u_1 r_0 r is a ring isomorphism R -> u_1R.</p>
<p>Conversely, let R*G=bigoplus_gin G A_g be an (intrinsic) crossed product and select for each gin G a unit u_gin A_g of R*G. This is called a <em>basis of units for the crossed product</em> R*G. Then the maps a:G -> Aut(R) and t:Gx G -> R^* given by</p>
<p class="pcenter">
r^g = u_g^{-1} r u_g, \quad t(g,h) = u_{gh}^{-1} u_g u_h \quad (g,h \in G, r \in R)
</p>
<p>satisfy conditions (1) and (2) and R*G = R*_a^t G.</p>
<p>The choice of a basis of units u_g in A_g determines the action a and twisting t. If u_g in A_g : g in G and v_g in A_g : g in G are two sets of units of R*G then v_g = u_g r_g for some units r_g of R. Changing the basis of units results in a change of the action and the twisting and so changes the extrinsic definition of the crossed product but it does not change the intrinsic crossed product.</p>
<p>It is customary to select u_1=1. In that case a(1) is the identity map of R and t(1,g)=t(g,1)=1 for each g in G.</p>
<p><a id="X828C42CD86AF605F" name="X828C42CD86AF605F"></a></p>
<h4>7.7 <span class="Heading">Cyclic Crossed Products</span></h4>
<p>Let R*G=bigoplus_g in G A_g be a <em>crossed product</em> (<a href="chap7.html#X7FB21779832CE1CB"><b>7.6</b></a>) and assume that G = < g > is cyclic. Then the crossed product can be given using a particularly nice description.</p>
<p>Select a unit u in A_g, and let a be the automorphism of R given by r^a = u^-1 r u.</p>
<p>If G is infinite then set u_g^k = u^k for every integer k. Then</p>
<p class="pcenter">
R*G = R[ u | ru = u r^a ],
</p>
<p>a skew polynomial ring. Therefore in this case R*G is determined by</p>
<p class="pcenter">
[ R, a ].
</p>
<p>If G is finite of order d then set u_g^k = u^k for 0 le k < d. Then b = u^d in R and</p>
<p class="pcenter">
R*G = R[ u | ru = u r^a, u^d = b ]
</p>
<p>Therefore, R*G is completely determined by the following data:</p>
<p class="pcenter">
[ R , [ d , a , b ] ]
</p>
<p><a id="X7869E2A48784C232" name="X7869E2A48784C232"></a></p>
<h4>7.8 <span class="Heading">Abelian Crossed Products</span></h4>
<p>Let R*G=bigoplus_g in G A_g be a <em>crossed product</em> (<a href="chap7.html#X7FB21779832CE1CB"><b>7.6</b></a>) and assume that G is abelian. Then the crossed product can be given using a simple description.</p>
<p>Express G as a direct sum of cyclic groups:</p>
<p class="pcenter">
G = \langle g_1 \rangle \times \cdots \times \langle g_n \rangle
</p>
<p>and for each i=1,dots,n select a unit u_i in A_g_i.</p>
<p>Each element g of G has a unique expression</p>
<p class="pcenter">
g = g_1^{k_1} \cdots g_n^{k_n},
</p>
<p>where k_i is an arbitrary integer, if g_i has infinite order, and 0 le k_i < d_i, if g_i has finite order d_i. Then one selects a basis for the crossed product by taking</p>
<p class="pcenter">
u_g = u_{g_1^{k_1} \cdots g_n^{k_n}} = u_1^{k_1} \cdots u_n^{k_n}.
</p>
<p>* For each i=1,dots, n, let a_i be the automorphism of R given by r^a_i = u_i^-1 r u_i.</p>
<p>* For each 1 le i < j le n, let t_i,j = u_j^-1 u_i^-1 u_j u_i in R.</p>
<p>* If g_i has finite order d_i, let b_i=u_i^d_i in R.</p>
<p>Then</p>
<p class="pcenter">
R*G = R[u_1,\dots,u_n | ru_i = u_i r^{a_i}, u_j u_i = t_{ij} u_i u_j, u_i^{d_i} = b_i (1 \le i < j \le n) ],
</p>
<p>where the last relation vanishes if g_i has infinite order.</p>
<p>Therefore R*G is completely determined by the following data:</p>
<p class="pcenter">
[ R , [ d_i , a_i , b_i ]_{i=1}^n, [ t_{i,j} ]_{1 \le i < j \le n} ].
</p>
<p><a id="X80BABE5078A29793" name="X80BABE5078A29793"></a></p>
<h4>7.9 <span class="Heading">Classical crossed products</span></h4>
<p>A <em>classical crossed product</em> is a crossed product L*_a^t G, where L/K is a finite Galois extension, G=Gal(L/K) is the Galois group of L/K and a is the natural action of G on L. Then t is a 2-cocycle and the <em>crossed product</em> (<a href="chap7.html#X7FB21779832CE1CB"><b>7.6</b></a>) L*_a^t G is denoted by (L/K,t). The crossed product (L/K,t) is known to be a central simple K-algebra <a href="chapBib.html#biBR">[Rei03]</a>.</p>
<p><a id="X84C98BB8859BBEE2" name="X84C98BB8859BBEE2"></a></p>
<h4>7.10 <span class="Heading">Cyclic Algebras</span></h4>
<p>A <em>cyclic algebra</em> is a <em>classical crossed product</em> (<a href="chap7.html#X80BABE5078A29793"><b>7.9</b></a>) (L/K,t) where L/K is a finite cyclic field extension. The cyclic algebras have a very simple form.</p>
<p>Assume that Gal(L/K) is generated by g and has order d. Let u=u_g be the basis unit (<a href="chap7.html#X7FB21779832CE1CB"><b>7.6</b></a>) of the crossed product corresponding to g and take the remaining basis units for the crossed product by setting u_g^i = u^i, (i = 0, 1, dots, d-1). Then a = u^n in K. The cyclic algebra is usually denoted by (L/K,a) and one has the following description of (L/K,t)</p>
<p class="pcenter">
(L/K,t) = (L/K,a) = L[u| r u = u r^g, u^d = a ].
</p>
<p><a id="X8099A8C784255672" name="X8099A8C784255672"></a></p>
<h4>7.11 <span class="Heading">Cyclotomic algebras</span></h4>
<p>A <em>cyclotomic algebra</em> over F is a <em>classical crossed product</em> (<a href="chap7.html#X80BABE5078A29793"><b>7.9</b></a>) (F(xi)/F,t), where F is a field, xi is a root of unity in an extension of F and t(g,h) is a root of unity for every g and h in Gal(F(xi)/F).</p>
<p>The <em>Brauer-Witt Theorem</em> <a href="chapBib.html#biBY">[Yam74]</a> asserts that every <em>Wedderburn component</em> (<a href="chap7.html#X87273420791F220E"><b>7.3</b></a>) of a group algebra is <em>Brauer equivalent</em> (<a href="chap7.html#X7A24D5407F72C633"><b>7.5</b></a>) (over its centre) to a cyclotomic algebra.</p>
<p><a id="X84A142407B7565E0" name="X84A142407B7565E0"></a></p>
<h4>7.12 <span class="Heading">Numerical description of cyclotomic algebras</span></h4>
<p>Let A=(F(xi)/F,t) be a <em>cyclotomic algebra</em> (<a href="chap7.html#X8099A8C784255672"><b>7.11</b></a>), where xi=xi_k is a k-th root of unity. Then the Galois group G=Gal(F(xi)/F) is abelian and therefore one can obtain a simplified form for the description of cyclotomic algebras as for any <em>abelian crossed product</em> (<a href="chap7.html#X7869E2A48784C232"><b>7.8</b></a>).</p>
<p>Then the n x n matrix algebra M_n(A) can be described numerically in one of the following forms:</p>
<p>* If F(xi)=F, (i.e. G=1) then A=M_n(F) and thus the only data needed to describe A are the matrix size n and the field F:</p>
<p class="pcenter">
[n,F]
</p>
<p>* If G is cyclic (but not trivial) of order d then A is a cyclic cyclotomic algebra</p>
<p class="pcenter">
A = F(\xi) [ u | \xi u = u \xi^\alpha, u^d = \xi^\beta ]
</p>
<p>and so M_n(A) can be described with the following data</p>
<p class="pcenter">
[n,F,k,[d,\alpha,\beta]],
</p>
<p>where the integers k, d, alpha and beta satisfy the following conditions:</p>
<p class="pcenter">
\alpha^d \equiv 1 \; mod \; k, \quad
\beta(\alpha-1) \equiv 0 \; mod \; k.
</p>
<p>* If G is abelian but not cyclic then M_n(A) can be described with the following data (see <a href="chap7.html#X7869E2A48784C232"><b>7.8</b></a>):</p>
<p class="pcenter">
[n,F,k,[d_i,\alpha_i,\beta_i]_{i=1}^m, [\gamma_{i,j}]_{1\le i < j \le m} ]
</p>
<p>representing the n x n matrix ring over the following algebra:</p>
<p class="pcenter">
A = F(\xi)[ u_1, \ldots, u_m \mid
\xi u_i = u_i \xi^{\alpha_i}, \quad
u_i^{d_i}=\xi^{\beta_i}, \quad
u_s u_r = \xi^{\gamma_{rs}} u_r u_s, \quad
i = 1, \ldots, m, \quad
0 \le r < s \le m ]
</p>
<p>where</p>
<p>* g_1,...,g_m is an independent set of generators of G,</p>
<p>* d_i is the order of g_i,</p>
<p>* alpha_i, beta_i and gamma_rs are integers, and</p>
<p class="pcenter">
\xi^{g_i} = \xi^{\alpha_i}.
</p>
<p><a id="X8310E96086509397" name="X8310E96086509397"></a></p>
<h4>7.13 <span class="Heading">Idempotents given by subgroups</span></h4>
<p>Let G be a finite group and F a field whose characteristic does not divide the order of G. If H is a subgroup of G then set</p>
<p class="pcenter">
\widehat{H} = |H|^{-1}\sum_{x \in H} x.
</p>
<p>The element widehatH is an idempotent of FG which is central in FG if and only if H is normal in G.</p>
<p>If H is a proper normal subgroup of a subgroup K of G then set</p>
<p class="pcenter"> \varepsilon(K,H) = \prod_{L} (\widehat{N}-\widehat{L})
</p>
<p>where L runs on the normal subgroups of K which are minimal among the normal subgroups of K containing N properly. By convention, varepsilon(K,K)=widehatK. The element varepsilon(K,H) is an idempotent of FG.</p>
<p>If H and K are subgroups of G such that H is normal in K then e(G,K,H) denotes the sum of all different G-conjugates of varepsilon(K,H). The element e(G,K,H) is central in FG. In general it is not an idempotent but if the different conjugates of varepsilon(K,H) are orthogonal then e(G,K,H) is a central idempotent of FG.</p>
<p>If (K,H) is a Shoda Pair (<a href="chap7.html#X80C058BE81824B23"><b>7.14</b></a>) of G then there is a non-zero rational number a such that ae(G,K,H)) is a <em>primitive central idempotent</em> (<a href="chap7.html#X87B6505C7C2EE054"><b>7.4</b></a>) of the rational group algebra ℚ G. If (K,H) is a strong Shoda pair (<a href="chap7.html#X81DAF5267D30C83A"><b>7.15</b></a>) of G then e(G,K,H) is a primitive central idempotent of ℚ G.</p>
<p>Assume now that F is a finite field of order q, (K,H) is a strong Shoda pair of G and C is a cyclotomic class of K/H containing a generator of K/H. Then e_C(G,K,H) is a primitive central idempotent of FG (see <a href="chap7.html#X800D8C5087D79DC8"><b>7.17</b></a>).</p>
<p><a id="X80C058BE81824B23" name="X80C058BE81824B23"></a></p>
<h4>7.14 <span class="Heading">Shoda pairs</span></h4>
<p>Let G be a finite group. A <em>Shoda pair</em> of G is a pair (K,H) of subgroups of G for which there is a linear character chi of K with kernel H such that the induced character chi^G in G is irreducible. By <a href="chapBib.html#biBS">[Sho33]</a> or <a href="chapBib.html#biBORS">[ORS04]</a>, (K,H) is a Shoda pair if and only if the following conditions hold:</p>
<p>* H is normal in K,</p>
<p>* K/H is cyclic and</p>
<p>* if K^g cap K subseteq H for some g in G then g in K.</p>
<p>If (K,H) is a Shoda pair and chi is a linear character of Kle G with kernel H then the <em>primitive central idempotent</em> (<a href="chap7.html#X87B6505C7C2EE054"><b>7.4</b></a>) of ℚ G associated to the irreducible character chi^G is of the form e=e_ℚ (chi^G)=a e(G,K,H) for some a in ℚ <a href="chapBib.html#biBORS">[ORS04]</a> (see <a href="chap7.html#X8310E96086509397"><b>7.13</b></a> for the definition of e(G,K,H)). In that case we say that e is the <em>primitive central idempotent realized by the Shoda pair</em> (K,H) of G.</p>
<p>A group G is monomial, that is every irreducible character of G is monomial, if and only if every primitive central idempotent of ℚ G is realizable by a Shoda pair of G.</p>
<p><a id="X81DAF5267D30C83A" name="X81DAF5267D30C83A"></a></p>
<h4>7.15 <span class="Heading">Strong Shoda pairs</span></h4>
<p>A <em>strong Shoda pair</em> of G is a pair (K,H) of subgroups of G satisfying the following conditions:</p>
<p>* H is normal in K and K is normal in the normalizer N of H in G,</p>
<p>* K/H is cyclic and a maximal abelian subgroup of N/H and</p>
<p>* for every g in G\ N , varepsilon(K,H)varepsilon(K,H)^g=0. (See <a href="chap7.html#X8310E96086509397"><b>7.13</b></a> for the definition of varepsilon(K,H)).</p>
<p>Let (K,H) be a strong Shoda pair of G. Then (K,H) is a Shoda pair (<a href="chap7.html#X80C058BE81824B23"><b>7.14</b></a>) of G. Thus there is a linear character theta of K with kernel H such that the induced character chi=chi(G,K,H)=theta^G is irreducible. Moreover the <em>primitive central idempotent</em> (<a href="chap7.html#X87B6505C7C2EE054"><b>7.4</b></a>) e_ℚ (chi) of ℚ G realized by (K,H) is e(G,K,H), see <a href="chapBib.html#biBORS">[ORS04]</a>.</p>
<p>Two <em>strong Shoda pairs</em> (<a href="chap7.html#X81DAF5267D30C83A"><b>7.15</b></a>) (K_1,H_1) and (K_2,H_2) of G are said to be <em>equivalent</em> if the characters chi(G,K_1,H_1) and chi(G,K_2,H_2) are Galois conjugate, or equivalently if e(G,K_1,H_1)=e(G,K_2,H_2).</p>
<p>The advantage of strong Shoda pairs over Shoda pairs is that one can describe the simple algebra FGe_F(chi) as a matrix algebra of a <em>cyclotomic algebra</em> (<a href="chap7.html#X8099A8C784255672"><b>7.11</b></a>, see <a href="chapBib.html#biBORS">[ORS04]</a> for F=ℚ and <a href="chapBib.html#biBO">[Olt07]</a> for the general case).</p>
<p>More precisely, ℚ Ge(G,K,H) is isomorphic to M_n(ℚ (xi)*_a^t N/K), where xi is a [K:H]-th root of unity, N is the normalizer of H in G, n=[G:N] and ℚ (xi)*_a^t N/K is a <em>crossed product</em> (see <a href="chap7.html#X7FB21779832CE1CB"><b>7.6</b></a>) with action a and twisting t given as follows:</p>
<p>Let x be a fixed generator of K/H and varphi : N/K -> N/H a fixed left inverse of the canonical projection N/H-> N/K. Then</p>
<p class="pcenter">
\xi^{a(r)} = \xi^i, \mbox{ if } x^{\varphi(r)}= x^i
</p>
<p>and</p>
<p class="pcenter">
t(r,s) = \xi^j, \mbox{ if } \varphi(rs)^{-1} \varphi(r)\varphi(s) = x^j,
</p>
<p>for r,s in N/K and integers i and j, see <a href="chapBib.html#biBORS">[ORS04]</a>. Notice that the cocycle is the one given by the natural extension</p>
<p class="pcenter">
1 \rightarrow K/H \rightarrow N/H \rightarrow N/K \rightarrow 1
</p>
<p>where K/H is identified with the multiplicative group generated by xi. Furthermore the centre of the algebra is ℚ (chi), the field of character values over ℚ, and N/K is isomorphic to Gal(ℚ (xi)/ℚ (chi)).</p>
<p>If the rational field is changed to an arbitrary ring F of characteristic 0 then the Wedderburn component A_F(chi), where chi = chi(G,K,H) is isomorphic to F(chi)otimes_ℚ (chi)A_ℚ (chi). Using the description given above of A_ℚ (chi)=ℚ G e(G,K,H) one can easily describe A_F(chi) as M_nd(F(xi)/F(chi),t'), where d=[ℚ (xi): ℚ(chi)]/[F(xi):F(chi)] and t' is the restriction to Gal(F(xi)/F(chi)) of t (a cocycle of N/K = Gal(ℚ (xi)/ℚ (chi))).</p>
<p><a id="X84C694978557EFE5" name="X84C694978557EFE5"></a></p>
<h4>7.16 <span class="Heading">Strongly monomial characters and strongly monomial groups</span></h4>
<p>Let G be a finite group an chi an irreducible character of G.</p>
<p>One says that chi is <em>strongly monomial</em> if there is a <em>strong Shoda pair</em> (<a href="chap7.html#X81DAF5267D30C83A"><b>7.15</b></a>) (K,H) of G and a linear character theta of K of G with kernel H such that chi=theta^G.</p>
<p>The group G is <em>strongly monomial</em> if every irreducible character of G is strongly monomial.</p>
<p>Strong Shoda pairs where firstly introduced by Olivieri, del Río and Simón who proved that every abelian-by-supersolvable group is strongly monomial <a href="chapBib.html#biBORS">[ORS04]</a>. The algorithm to compute the Wedderburn decomposition of rational group algebras for strongly monomial groups was explained in <a href="chapBib.html#biBOR">[OR03]</a>. This method was extended for semisimple finite group algebras by Broche Cristo and del Río in <a href="chapBib.html#biBBR">[BR07]</a> (see Section <a href="chap7.html#X800D8C5087D79DC8"><b>7.17</b></a>). Finally, Olteanu <a href="chapBib.html#biBO">[Olt07]</a> shows how to compute the <em>Wedderburn decomposition</em> (<a href="chap7.html#X87273420791F220E"><b>7.3</b></a>) of an arbitrary semisimple group ring by making use of not only the strong Shoda pairs of G but also the strong Shoda pairs of the subgroups of G.</p>
<p>.</p>
<p><a id="X800D8C5087D79DC8" name="X800D8C5087D79DC8"></a></p>
<h4>7.17 <span class="Heading">Cyclotomic Classes and Strong Shoda Pairs</span></h4>
<p>Let G be a finite group and F a finite field of order q, coprime to the order of G.</p>
<p>Given a positive integer n, coprime to q, the q-<em>cyclotomic classes</em> modulo n are the set of residue classes module n of the form</p>
<p class="pcenter">
\{i,iq,iq^2,iq^3, \dots \}
</p>
<p>The q-cyclotomic classes module n form a partition of the set of residue classes module n.</p>
<p>A <em>generating cyclotomic class </em> module n is a cyclotomic class containing a generator of the additive group of residue classes module n, or equivalently formed by integers coprime to n.</p>
<p>Let (K,H) be a strong Shoda pair (<a href="chap7.html#X81DAF5267D30C83A"><b>7.15</b></a>) of G and set n=[K:H]. Fix a primitive n-th root of unity xi in some extension of F and an element g of K such that gH is a generator of K/H. Let C be a generating q-cyclotomic class modulo n. Then set</p>
<p class="pcenter">
\varepsilon_C(K,H) = [K:H]^{-1} \widehat{H} \sum_{i=0}^{n-1} tr(\xi^{-ci})g^i,
</p>
<p>where c is an arbitrary element of C and tr is the trace map of the field extension F(xi)/F. Then varepsilon_C(K,H) does not depend on the choice of c in C and is a <em>primitive central idempotent</em> (<a href="chap7.html#X87B6505C7C2EE054"><b>7.4</b></a>) of FK.</p>
<p>Finally, let e_C(G,K,H) denote the sum of the different G-conjugates of varepsilon_C(K,H). Then e_C(G,K,H) is a <em>primitive central idempotent</em> (<a href="chap7.html#X87B6505C7C2EE054"><b>7.4</b></a>) of FG <a href="chapBib.html#biBBR">[BR07]</a>. We say that e_C(G,K,H) is the primitive central idempotent realized by the strong Shoda pair (K,H) of the group G and the cyclotomic class C.</p>
<p>If G is <em>strongly monomial</em> (<a href="chap7.html#X84C694978557EFE5"><b>7.16</b></a>) then every primitive central idempotent of FG is realizable by some <em>strong Shoda pair</em> (<a href="chap7.html#X81DAF5267D30C83A"><b>7.15</b></a>) of G and some cyclotomic class C <a href="chapBib.html#biBBR">[BR07]</a>. As in the zero characteristic case, this explain how to compute the <em>Wedderburn decomposition</em> (<a href="chap7.html#X87273420791F220E"><b>7.3</b></a>) of FG for a finite semisimple algebra of a strongly monomial group (see <a href="chapBib.html#biBBR">[BR07]</a> for details). For non strongly monomial groups the algorithm to compute the Wedderburn decomposition just uses the Brauer characters.</p>
<div class="chlinkprevnextbot"> <a href="chap0.html">Top of Book</a> <a href="chap6.html">Previous Chapter</a> <a href="chapBib.html">Next Chapter</a> </div>
<div class="chlinkbot"><span class="chlink1">Goto Chapter: </span><a href="chap0.html">Top</a> <a href="chap1.html">1</a> <a href="chap2.html">2</a> <a href="chap3.html">3</a> <a href="chap4.html">4</a> <a href="chap5.html">5</a> <a href="chap6.html">6</a> <a href="chap7.html">7</a> <a href="chapBib.html">Bib</a> <a href="chapInd.html">Ind</a> </div>
<hr />
<p class="foot">generated by <a href="http://www.math.rwth-aachen.de/~Frank.Luebeck/GAPDoc">GAPDoc2HTML</a></p>
</body>
</html>
distrib
>
Mandriva
>
2010.0
>
i586
>
media
>
contrib-release
>
by-pkgid
>
5e1854624d3bc613bdd0dd13d1ef9ac7
>
files
>
3364
