betaFeedback

WorksMathematics

The conductors of fibered Burnside rings

Abstract

We give an explicit formula for the conductors of the primitive idempotents of fibered Burnside algebras of finite rank. This formula generalizes previous results for Burnside rings, monomial representation rings, and cyclic fibers, removing the coprimality condition required in the latter case. We also derive some consequences for cyclic fiber groups and pp-groups.

Introduction

The conductors of the primitive idempotents of the rational Burnside algebra were first determined by Nicolson in 1978 [8]. In his article, Nicolson showed that, for a primitive idempotent eHB(G)e_{H}^{B(G)} of BQ(G)B_{\mathbb{Q}}(G) (indexed by the conjugacy class of H≤GH \leq G), the least positive integer tt such that teHB(G)te_{H}^{B(G)} belongs to B(G)B(G) is precisely [NG(H):H][H:H′]0[N_G(H):H][H:H']_0, where [H:H′]0[H:H']_0 denotes the product of the prime divisors of [H:H′][H:H']. Later, Kratzer & Thévenaz [6] and Dress & Vallejo [3] provided more direct proofs of this result.

In the case of the monomial representation algebra DK(G)D_K(G), the primitive idempotents eH,hD(G)e_{H,h}^{D(G)} are indexed by the conjugacy classes of pairs (H,hH′)(H,hH') with H≤GH \leq G and h∈Hh \in H, where KK is a field of characteristic zero containing a primitive exp⁡(G)\exp(G)-th root of unity. Müller showed in [7 §3] that the conductor of eH,hD(G)e_{H,h}^{D(G)} is equal to the index [NG(H,hH′):H′][N_G(H,hH'):H']. In general, [NG(H,hH′):H′]eH,hD(G)[N_G(H,hH'):H']e_{H,h}^{D(G)} does not lie in D(G)D(G) but in DO(G)D_O(G), where OO is the ring of integers of KK.

A first generalization of the previous results in the context of fibered Burnside rings was given by Raggi-Cárdenas and the author in [4 §5]. For a positive integer nn, the primitive idempotents eH,hG,ne^{G,n}_{H,h} of BKn(G)B^n_{\mathbb{K}}(G), where K\mathbb{K} is a field of characteristic zero containing a primitive nn-th root of 11, are indexed by the conjugacy classes of pairs (H,hH[n])(H,hH^{[n]}), with H≤GH \leq G and h∈Hh \in H. Here, H[n]H^{[n]} is the minimal normal subgroup of HH such that H/H[n]H/H^{[n]} is an abelian group of exponent dividing nn [4 §1]. When nn is a divisor of exp⁡(G)\exp(G) such that gcd⁡(n,exp⁡(G)/n)=1\gcd(n,\exp(G)/n) = 1, the conductor of eH,hG,ne^{G,n}_{H,h} is precisely [NG(H,hH[n]):H[n]][H[n]:H′]0[N_G(H,hH^{[n]}) : H^{[n]}][H^{[n]} : H']_0. Although this result has restrictive hypotheses, it covers the cases of the Burnside ring (n=1n = 1) and the monomial representation ring (n=exp⁡(G)n = \exp(G)).

In this article, we focus on fibered Burnside rings BA(G)B^A(G) of finite rank. As shown by Boltje and Yılmaz in [2 §3], the primitive idempotents eH,ΦG,Ae^{G,A}_{H,\Phi} of the fibered Burnside algebra BKA(G)B^A_{\mathbb{K}}(G) over a sufficiently large field of characteristic zero are indexed by orbits of pairs (H,Φ)(H,\Phi) under a natural action of GG on the set DGAD^A_G. The main result provides an explicit formula for the conductor of eH,ΦG,Ae^{G,A}_{H,\Phi}, namely,

cH,ΦG,A=[NG(H,Φ):H]∣Hom⁡(H,A)∣rA(H),c^{G,A}_{H,\Phi} = [N_G(H,\Phi) : H] |\operatorname{Hom}(H,A)| r^A(H),

where rA(H)r^A(H) is an additional factor depending only on HH and AA. As in the previous cases, Boltje’s integrality criterion and the Möbius function play an important role in the proof. A closer examination of the values of the Möbius function allows us to remove the coprimality condition imposed in [4].

The article is organized into two sections. The first collects the necessary preliminaries on abelian groups, the Möbius function, and fibered Burnside rings. The second is devoted to the statement and proof of the main result, together with some consequences concerning cyclic fibers and pp-groups.

Notation

For a positive integer nn, we write n0n_0 for its largest square-free divisor, that is, the product of its distinct prime divisors. We write SG\mathcal{S}_G for the poset of subgroups of a group GG, and [SG][\mathcal{S}_G] for a subset of representatives of the conjugacy classes of SG\mathcal{S}_G. If XX is a GG-set, then X/GX/G denotes its set of orbits, and we write x=Gyx =_G y for elements x,y∈Xx,y \in X lying in the same orbit. We regard GG and SG\mathcal{S}_G as GG-sets under conjugation, writing gx{}^{g}x instead of gxg−1gxg^{-1}. For H≤GH \leq G and an abelian group AA, ϕ∈Hom⁡(H,A)\phi\in\operatorname{Hom}(H,A), and g∈Gg \in G, we write gϕ{}^{g}\phi for the morphism defined by (gϕ)(x)=ϕ(g−1xg)({}^{g}\phi)(x) = \phi(g^{-1}xg), for x∈gHx \in{}^{g}H. For K≤HK \leq H, we denote by ϕ∣K\phi|_K the restriction of ϕ\phi to KK. All rings RR are assumed to be associative, unital, and commutative, and we use the standard notation Z\mathbb{Z}, Q\mathbb{Q}, R\mathbb{R}, and C\mathbb{C} for the usual number systems.

Preliminaries

We recall some facts about morphisms to abelian groups and the Möbius function of a poset of subgroups, and then review the basic properties of fibered Burnside rings that will be used throughout the paper.

Morphisms to abelian groups

For an abelian group AA and a positive integer nn, we denote by tor⁡n(A)\operatorname{tor}_{n}(A) the subgroup of elements of AA whose orders divide nn; we say that AA is an nn-torsion abelian group if tor⁡n(A)=A\operatorname{tor}_{n}(A)=A. For a finite group GG, the subgroup G[n]G^{[n]} is defined as the smallest normal subgroup T⊴GT\trianglelefteq G such that G/TG/T is an nn-torsion abelian group; it can be constructed as the subgroup generated by G′G' ∪\cup {gn∣g∈G}\{g^{n}\mid g\in G\}. By [4 Lemma 1.1], G[exp⁡(G)]=G′G^{[\exp(G)]}=G', G[1]=GG^{[1]}=G, H[n]≤G[n]H^{[n]}\leq G^{[n]} for H≤GH\leq G, and G[n]≤G[m]G^{[n]}\leq G^{[m]} if m∣nm\mid n. Also, there is a natural isomorphism

Hom⁡(G,tor⁡n(A))≅Hom⁡(G/G[n],A)\operatorname{Hom}(G,\operatorname{tor}_{n}(A))\cong\operatorname{Hom}(G/G^{[n]},A)

by [4 Lemma 1.2]. In particular,

Hom⁡(G,A)≅Hom⁡(G,tor⁡exp⁡(G)(A))≅Hom⁡(G/G[n(A,G)],A),\operatorname{Hom}(G,A)\cong\operatorname{Hom}(G,\operatorname{tor}_{\exp(G)}(A))\cong\operatorname{Hom}(G/G^{[n(A,G)]},A),

where n(A,G):=exp⁡(tor⁡exp⁡(G)(A))n(A,G):=\exp(\operatorname{tor}_{\exp(G)}(A)). By [4 Lemma 1.3], Hom⁡(G,Cn)\operatorname{Hom}(G,C_n) and G/G[n]G/G^{[n]} are isomorphic, and so ∣Hom⁡(G,Cn)∣=[G:G[n]]|\operatorname{Hom}(G,C_n)|=[G:G^{[n]}].

Lemma 1.1. Let nn be a positive integer and GG be a finite abelian group. Let pp be a prime divisor of ∣G[n]∣|G^{[n]}|.

  1. If pp divides [G:G[n]][G:G^{[n]}], then GG has a pp-element of order greater than pp;

  2. if pp does not divide [G:G[n]][G:G^{[n]}], then pp does not divide nn.

Proof. Since GG is finite and abelian, the map fn:G→Gf_n:G\to G given by g↦gng\mapsto g^n is an endomorphism with Im⁡(fn)=G[n]\operatorname{Im}(f_n)=G^{[n]}, and ∣ker⁡(fn)∣=[G:G[n]]|\ker(f_n)|=[G:G^{[n]}]. Because p∣∣G[n]∣p\mid|G^{[n]}|, the pp-primary component GpG_p is non-trivial, and fnf_n restricts to an endomorphism on GpG_p with image Gp[n]=G[n]∩Gp≠1G_p^{[n]}=G^{[n]}\cap G_p\neq1. If p∣[G:G[n]]p\mid[G:G^{[n]}], then pp divides exp⁡(G/G[n])\exp(G/G^{[n]}), which divides nn. Then there is some g∈Gpg\in G_p such that 1≠gn∈Gp[n]1\neq g^n\in G_p^{[n]}; since p∣np\mid n, it follows that gg has order greater than pp. On the other hand, if p∤[G:G[n]]p\nmid[G:G^{[n]}], then Gp∩ker⁡(fn)=1G_p\cap\ker(f_n)=1, hence fnf_n is an automorphism on GpG_p and so nn is coprime to pp.

If AA is a finite abelian group, we consider a decomposition A=Cd1×Cd2×⋯×CdrA=C_{d_1}\times C_{d_2}\times\cdots\times C_{d_r} into cyclic factors. For any finite group GG, there is a natural isomorphism Hom⁡(G,A)≅∏i=1rHom⁡(G,Cdi)\operatorname{Hom}(G,A)\cong\prod_{i=1}^{r}\operatorname{Hom}(G,C_{d_i}), and for each H≤GH\leq G, the following diagram involving the restriction maps commutes:

Hom⁡(G,A)→∏i=1rHom⁡(G,Cdi)res⁡HG↓↓∏ires⁡HGHom⁡(H,A)→∏i=1rHom⁡(H,Cdi).\begin{CD} \operatorname{Hom}(G,A) @>>> \prod_{i=1}^{r}\operatorname{Hom}(G,C_{d_i}) \\ @V{\operatorname{res}^{G}_{H}}VV @VV{\prod_i\operatorname{res}^{G}_{H}}V \\ \operatorname{Hom}(H,A) @>>> \prod_{i=1}^{r}\operatorname{Hom}(H,C_{d_i}). \end{CD}

Indeed, for every ϕ:G⟶A\phi:G\longrightarrow A and all 1≤i≤r1\leq i\leq r, πCdi∘(ϕ∣H)=(πCdi∘ϕ)∣H\pi_{C_{d_i}}\circ(\phi|_H)=(\pi_{C_{d_i}}\circ\phi)|_H.

Lemma 1.2. Let H≤GH\leq G, and consider the restriction map res⁡HG:Hom⁡(G,A)⟶Hom⁡(H,A)\operatorname{res}^{G}_{H}:\operatorname{Hom}(G,A)\longrightarrow\operatorname{Hom}(H,A). The following holds:

  1. For ψ∈Im⁡(res⁡HG)\psi\in\operatorname{Im}(\operatorname{res}^{G}_{H}),

∣(res⁡HG)−1(ψ)∣=∏i=1r[G:HG[di]];\left|(\operatorname{res}^{G}_{H})^{-1}(\psi)\right|=\prod_{i=1}^{r}[G:HG^{[d_i]}];
  1. Let Φ:Hom⁡(G,A)⟶K×\Phi:\operatorname{Hom}(G,A)\longrightarrow\mathbb{K}^{\times} be a linear K\mathbb{K}-character, where K\mathbb{K} is a field of characteristic zero. Then

∑ϕ∈(res⁡HG)−1(ψ)Φ(ϕ)={(∏i=1r[G:HG[di]])Φ(ϕ0)if ker⁡(res⁡HG)≤ker⁡Φ,0otherwise,\sum_{\phi\in(\operatorname{res}^{G}_{H})^{-1}(\psi)}\Phi(\phi)= \begin{cases} \left(\prod_{i=1}^{r}[G:HG^{[d_i]}]\right)\Phi(\phi_0)&\text{if }\ker(\operatorname{res}^{G}_{H})\leq\ker\Phi,\\ 0&\text{otherwise,} \end{cases}

where ϕ0\phi_0 is any element in (res⁡HG)−1(ψ)(\operatorname{res}^{G}_{H})^{-1}(\psi).

Proof. Note that (res⁡HG)−1(ψ)=ker⁡(res⁡HG)ϕ0(\operatorname{res}^{G}_{H})^{-1}(\psi)=\ker(\operatorname{res}^{G}_{H})\phi_0 for some ϕ0:G⟶A\phi_0:G\longrightarrow A such that ϕ0∣H=ψ\phi_0|_H=\psi, and that

ker⁡(Hom⁡(G,A)⟶Hom⁡(H,A))≅∏i=1rker⁡(Hom⁡(G,Cdi)⟶Hom⁡(H,Cdi))\ker(\operatorname{Hom}(G,A)\longrightarrow\operatorname{Hom}(H,A))\cong\prod_{i=1}^{r}\ker(\operatorname{Hom}(G,C_{d_i})\longrightarrow\operatorname{Hom}(H,C_{d_i}))

by the previous diagram. By [4 Lemma 5.3],

∣ker⁡(Hom⁡(G,Cdi)⟶Hom⁡(H,Cdi))∣=[G:HG[di]]\left|\ker(\operatorname{Hom}(G,C_{d_i})\longrightarrow\operatorname{Hom}(H,C_{d_i}))\right|=[G:HG^{[d_i]}]

for each ii, proving assertion 1. For the second assertion, we have

∑ϕ∈(res⁡HG)−1(ψ)Φ(ϕ)=∑λ∈ker⁡(res⁡HG)Φ(λϕ0)=(∑λ∈ker⁡(res⁡HG)Φ(λ))Φ(ϕ0)={∣ker⁡(res⁡HG)∣Φ(ϕ0)if ker⁡(res⁡HG)≤ker⁡Φ,0otherwise,\begin{aligned} \sum_{\phi\in(\operatorname{res}^{G}_{H})^{-1}(\psi)}\Phi(\phi)=\sum_{\lambda\in\ker(\operatorname{res}^{G}_{H})}\Phi(\lambda\phi_0) \\ &=\left(\sum_{\lambda\in\ker(\operatorname{res}^{G}_{H})}\Phi(\lambda)\right)\Phi(\phi_0) \\ &=\begin{cases} \left|\ker(\operatorname{res}^{G}_{H})\right|\Phi(\phi_0)&\text{if }\ker(\operatorname{res}^{G}_{H})\leq\ker\Phi,\\ 0&\text{otherwise,} \end{cases} \end{aligned}

by orthogonality relations for ker⁡(res⁡HG)\ker(\operatorname{res}^{G}_{H}). Together with assertion 1, this proves 2. □\square

The Möbius function of the poset of subgroups

We consider SG\mathcal{S}_G as a poset with respect to inclusion. The Möbius function μ:SG×SG⟶Z\mu: \mathcal{S}_G \times\mathcal{S}_G \longrightarrow\mathbb{Z} is defined recursively as follows: for H≤GH \leq G, μ(H,H)=1\mu(H,H)=1, and for K≠HK \ne H,

μ(K,H)=−∑K<T≤Hμ(K,T).\mu(K,H)=-\sum_{K<T\leq H}\mu(K,T).

Equivalently, μ\mu is defined as the inverse of the zeta function in the incidence algebra of SG\mathcal{S}_G. It is easy to see that μ(K,H)=−1\mu(K,H)=-1 if KK is a maximal subgroup of HH. We write μ(G)\mu(G) instead of μ(1,G)\mu(1,G). Moreover, if N⊴GN \trianglelefteq G, then μ(N,G)=μ(G/N)\mu(N,G)=\mu(G/N) since SG/N\mathcal{S}_{G/N} and {T∈SG∣N≤T}\{T\in\mathcal{S}_G\mid N\leq T\} are isomorphic as posets. The following proposition collects some properties of μ\mu proved by Hawkes, Isaacs and Özaydin in [5].

Proposition 1.3 ([5]). Let GG be a finite group.

  1. If N⊴GN\trianglelefteq G, let K(N,G)\mathcal{K}(N,G) denote the set of complements of NN in GG. Then

μ(G)=μ(G/N)(∑X∈K(N,G)μ(X,G)).\mu(G)=\mu(G/N)\left(\sum_{X\in\mathcal{K}(N,G)}\mu(X,G)\right).
  1. If GG is a pp-group of order prp^r, then μ(G)≠0\mu(G)\ne0 if and only if GG is an elementary abelian pp-group. In this case, μ(G)=(−1)rp(r2)\mu(G)=(-1)^r p^{\binom{r}{2}}.

  1. If H≤GH\leq G, then [NG(H):H][N_G(H):H] divides μ(H,G)[G:HG′]0\mu(H,G)[G:HG']_0.

Next, we establish a vanishing criterion for the Möbius function in terms of the indices determined by the subgroups HG[n]HG^{[n]}. This will be useful in the proof of our main result.

Lemma 1.4. Let H⊴GH\trianglelefteq G and nn be a positive integer. If [G:HG[n]][G:HG^{[n]}] and [HG[n]:HG′][HG^{[n]}:HG'] are not coprime, then μ(H,G)=0\mu(H,G)=0.

Proof. The quotient G/HG′G/HG' is abelian and (G/HG′)[n]=HG[n]/HG′(G/HG')^{[n]}=HG^{[n]}/HG'. By part 1 of Proposition 1.3, we have μ(H,G)=μ(HG′,G)⋅T\mu(H,G)=\mu(HG',G)\cdot T for some integer TT. Since [G:HG[n]][G:HG^{[n]}] and [HG[n]:HG′][HG^{[n]}:HG'] have a common prime divisor pp, it follows from Lemma 1.1 that G/HG′G/HG' has a pp-element of order greater than pp. Consequently, there is a subgroup HG′≤M≤GHG'\leq M\leq G such that G/MG/M is cyclic of order p2p^2. Applying part 1 of Proposition 1.3 again, μ(HG′,G)=μ(M,G)⋅U\mu(HG',G)=\mu(M,G)\cdot U for some integer UU. Finally, μ(M,G)=μ(G/M)=0\mu(M,G)=\mu(G/M)=0 by part 2 of Proposition 1.3, and the result follows. □

Fibered Burnside rings

Given a finite group GG and an abelian group AA, an AA-fibered GG-set is a G×AG\times A-set on which AA acts freely and with finitely many orbits. The category of AA-fibered GG-sets and G×AG \times A-equivariant functions, denoted by gsetA\mathrm{gset}^{A}, is a symmetric semiring category for disjoint unions and a tensor product ⊗A\otimes_{A}. The AA-fibered Burnside ring of GG is defined as the Grothendieck ring of the category gsetA\mathrm{gset}^{A}, denoted by BA(G)B^{A}(G).

The following presentation of BA(G)B^{A}(G) is often more convenient for computations: let MGA\mathcal{M}^{A}_{G} be the set of AA-monomial pairs or AA-subcharacters of GG, that is, pairs of the form (H,ϕ)(H,\phi) where H≤GH \leq G and ϕ∈Hom⁡(H,A)\phi\in\operatorname{Hom}(H,A); the group GG acts on the set MGA\mathcal{M}^{A}_{G} by conjugation, and the stabilizer and orbit of a pair (H,ϕ)(H,\phi) are denoted by NG(H,ϕ)N_{G}(H,\phi) and [H,ϕ]G[H,\phi]_{G}, respectively. As the transitive AA-fibered GG-sets are in one-to-one correspondence with the orbits of MGA\mathcal{M}^{A}_{G}, the ring BA(G)B^{A}(G) is the free abelian group on MGA/G\mathcal{M}^{A}_{G}/G, with multiplication determined on the basis by

[H,ϕ]G[K,ψ]G=∑HgK∈H\G/K[H∩gK,ϕ∣H∩gKgψ∣H∩gK]G.[H,\phi]_{G}[K,\psi]_{G}=\sum_{HgK\in H\backslash G/K}[H\cap{}^{g}K,\phi|_{H\cap{}^{g}K}{}^{g}\psi|_{H\cap{}^{g}K}]_{G}.

and with [G,1]G[G,1]_{G} acting as multiplicative identity; MGA/G\mathcal{M}^{A}_{G}/G is called the monomial basis of BA(G)B^{A}(G). It is known that BA(G)≅Btor⁡exp⁡(G)(A)(G)B^{A}(G)\cong B^{\operatorname{tor}_{\exp(G)}(A)}(G), and moreover, BA(G)B^{A}(G) has finite rank if and only if tor⁡exp⁡(G)(A)\operatorname{tor}_{\exp(G)}(A) is finite. Throughout this paper, we assume that tor⁡exp⁡(G)(A)\operatorname{tor}_{\exp(G)}(A) is finite.

For each H≤GH\leq G, there is a restriction map res⁡HG:BA(G)⟶BA(H)\operatorname{res}^{G}_{H}:B^{A}(G)\longrightarrow B^{A}(H) defined on the basis by

res⁡HG([K,ψ]G)=∑HgK∈H\G/K[H∩gK,gψ∣H∩gK]H.\operatorname{res}^{G}_{H}([K,\psi]_{G})=\sum_{HgK\in H\backslash G/K}[H\cap{}^{g}K,{}^{g}\psi|_{H\cap{}^{g}K}]_{H}.

as well as an induction map ind⁡HG:BA(H)⟶BA(G)\operatorname{ind}^{G}_{H}:B^{A}(H)\longrightarrow B^{A}(G), [K,ψ]H↦[K,ψ]G[K,\psi]_{H}\mapsto[K,\psi]_{G}. For H≤GH\leq G and g∈Gg\in G, there is also a gg-conjugation map cg,H:BA(H)⟶BA(gH)c_{g,H}:B^{A}(H)\longrightarrow B^{A}({}^{g}H), given by [K,ψ]H↦g[K,ψ]H=[gK,gψ]gH[K,\psi]_{H}\mapsto{}^{g}[K,\psi]_{H}=[{}^{g}K,{}^{g}\psi]_{{}^{g}H}. The collection of the rings BA(H)B^{A}(H) for H≤GH\leq G together with induction, restriction and conjugation maps is a Green functor over GG in the usual sense.

The ordinary Burnside ring B(G)B(G) embeds naturally into BA(G)B^{A}(G) via the ring homomorphism [G/H]↦[H,1]G[G/H]\mapsto[H,1]_{G}, admitting a retraction [H,ϕ]G↦[G/H][H,\phi]_{G}\mapsto[G/H]. There is also a surjective ring homomorphism πG:BA(G)⟶ZHom⁡(G,A)\pi_{G}:B^{A}(G)\longrightarrow\mathbb{Z}\operatorname{Hom}(G,A) mapping [K,ϕ]G[K,\phi]_{G} to ϕ\phi if K=GK=G, and to 00 otherwise; this map admits a section ϕ↦[G,ϕ]G\phi\mapsto[G,\phi]_{G}.

The group GG acts on the product of group rings ∏H≤GZHom⁡(H,A)\prod_{H\leq G}\mathbb{Z}\operatorname{Hom}(H,A) through the conjugation maps, and the ghost ring of BA(G)B^{A}(G) is defined as the ring of fixed points

B~A(G):=(∏H≤GZHom⁡(H,A))G.\widetilde{B}^{A}(G):=\left(\prod_{H\leq G}\mathbb{Z}\operatorname{Hom}(H,A)\right)^{G}.

The product map ρ=∏H≤GπH∘res⁡HG\rho=\prod_{H\leq G}\pi_{H}\circ\operatorname{res}^{G}_{H} is known as the mark morphism and embeds BA(G)B^{A}(G) into its ghost ring. It is explicitly given on the basis by

ρ([K,ψ]G)=(∑gK∈G/K,H≤gKgψ∣H)H≤G.\rho([K,\psi]_{G})=\left(\sum_{\substack{gK\in G/K,\\ H\leq{}^{g}K}}{}^{g}\psi|_{H}\right)_{H\leq G}.

We write BRA(G)B^{A}_{R}(G) for the scalar extension of BA(G)B^{A}(G) to a commutative ring RR, regarded as an RR-algebra, and B~RA(G)\widetilde{B}^{A}_{R}(G) for the corresponding extension of its ghost ring. The extensions of the maps defined in this section are denoted the same. Boltje’s Integrality Criterion [1 Theorem 2.2] characterizes the image of the extended mark morphism in B~RA(G)\widetilde{B}^{A}_{R}(G) as follows. Let

y=(∑ϕ∈Hom⁡(H,A)aH,ϕϕ)H≤Gy=\left(\sum_{\phi\in\operatorname{Hom}(H,A)}a_{H,\phi}\phi\right)_{H\leq G}

be an element of B~RA(G)\widetilde{B}^{A}_{R}(G). Then yy belongs to ρ(BRA(G))\rho\left(B^{A}_{R}(G)\right) if and only if, for every (K,ψ)∈MGA(K,\psi)\in\mathcal{M}^{A}_{G}, the congruence

∑(H,ϕ)∈MGA(K,ψ)≤(H,ϕ)μ(K,H)aH,ϕ≡0(mod[NG(K,ψ):K])(1)\sum_{\substack{(H,\phi)\in\mathcal{M}^{A}_{G}\\(K,\psi)\leq(H,\phi)}}\mu(K,H)a_{H,\phi}\equiv0\pmod{[N_{G}(K,\psi):K]} \tag*{(1)}

holds. Here, (K,ψ)≤(H,ϕ)(K,\psi)\leq(H,\phi) means that K≤HK\leq H and ψ=ϕ∣K\psi=\phi|_{K}. This defines a partial order on MGA\mathcal{M}^{A}_{G} respecting the action of GG.

Let K\mathbb{K} be a field of characteristic zero that is a splitting field for every Hom⁡(H,A)\operatorname{Hom}(H,A), with H≤GH\leq G. As noted in [2 §3], this is equivalent to requiring that K\mathbb{K} contains a root of unity of order exp⁡(tor⁡exp⁡(G)(A))\exp\left(\operatorname{tor}_{\exp(G)}(A)\right). We consider the set DGA\mathcal{D}^{A}_{G} consisting of pairs (H,Φ)(H,\Phi) where H≤GH\leq G and Φ∈Hom⁡(Hom⁡(H,A),K×)\Phi\in\operatorname{Hom}(\operatorname{Hom}(H,A),\mathbb{K}^{\times}); again, GG acts on this set by conjugation and the stabilizer and orbit of a pair (H,Φ)(H,\Phi) are denoted by NG(H,Φ)N_{G}(H,\Phi) and [H,Φ]G[H,\Phi]_{G}, respectively. If A=CnA=C_{n}, then DGCn\mathcal{D}^{C_{n}}_{G} can be identified with the set of pairs (H,hH[n])(H,hH^{[n]}), H≤GH\leq G and h∈Hh\in H. For each (H,Φ)(H,\Phi), the composite sH,ΦG,A:=Φ∘πH∘res⁡HGs^{G,A}_{H,\Phi}:=\Phi\circ\pi_{H}\circ\operatorname{res}^{G}_{H} is a ring homomorphism from BA(G)B^{A}(G) to K\mathbb{K}. By [2], two of such maps sH,ΦG,As^{G,A}_{H,\Phi} and sK,ΨG,As^{G,A}_{K,\Psi} are equal if and only if (H,Φ)=G(K,Ψ)(H,\Phi) =_{G} (K,\Psi). Moreover, these maps account for all elements of Hom⁡ring(BA(G),K)\operatorname{Hom}_{\mathrm{ring}}(B^{A}(G),\mathbb{K}), and BKA(G)B^{A}_{\mathbb{K}}(G) and ∏[H,Φ]G∈DGA/GK\prod_{[H,\Phi]_{G}\in\mathcal{D}^{A}_{G}/G}\mathbb{K} are isomorphic as K\mathbb{K}-algebras.

The primitive idempotents of BKA(G)B^{A}_{\mathbb{K}}(G) are in bijection with the orbits of DGA\mathcal{D}^{A}_{G}. More precisely, for each (H,Φ)∈DGA(H,\Phi)\in\mathcal{D}^{A}_{G} there is a primitive idempotent eH,ΦG,Ae^{G,A}_{H,\Phi} determined by

sK,ΨG,A(eH,ΦG,A)={1if (K,Ψ)=G(H,Φ),0else.s^{G,A}_{K,\Psi}\left(e^{G,A}_{H,\Phi}\right)= \begin{cases} 1 & \text{if }(K,\Psi)=_{G}(H,\Phi),\\ 0 & \text{else.} \end{cases}

In terms of the standard basis,

eH,ΦG,A=1∣NG(H,Φ)∣∣Hom⁡(H,A)∣∑K≤H,ϕ∈Hom⁡(H,A)∣K∣μ(K,H)Φ(ϕ−1)[K,ϕ∣K]G(2)e^{G,A}_{H,\Phi}= \frac{1}{\lvert N_{G}(H,\Phi)\rvert\lvert\operatorname{Hom}(H,A)\rvert} \sum_{\substack{K\leq H,\\ \phi\in\operatorname{Hom}(H,A)}} \lvert K\rvert\mu(K,H)\Phi(\phi^{-1})[K,\phi|_{K}]_{G} \tag*{(2)}

as showed in [2]. If O\mathcal{O} is the ring of algebraic integers of K\mathbb{K}, the conductor of a primitive idempotent eH,ΦG,Ae^{G,A}_{H,\Phi} is the least positive integer tt such that teH,ΦG,At e^{G,A}_{H,\Phi} belongs to BOA(G)B^{A}_{\mathcal{O}}(G), denoted by cH,ΦG,Ac^{G,A}_{H,\Phi}. In the case of cyclic fiber, we write eH,hG,Cne^{G,C_{n}}_{H,h} instead of eH,eνhG,Cne^{G,C_{n}}_{H,e\nu_{h}}, and CH,hG,nC^{G,n}_{H,h} for its conductor. The conductors do not depend on the choice of K\mathbb{K}, since the coefficients of the primitive idempotents with respect to the monomial basis are algebraic numbers.

As in [7], the behavior of certain idempotents with respect to induction and restriction will play a role in the proof of the main result. The following lemma collects the required properties. The proof is straightforward and follows the same ideas as the corresponding results in [4] and [7], so we omit it.

Lemma 1.5. Let (H,Φ)∈DGA(H,\Phi) \in D_G^A. Then:

res⁡HG(eH,ΦG,A)=∑Ψ∈Hom⁡(Hom⁡(H,A),K×)[H,Ψ]G=[H,Φ]GeH,ΨH,A.\operatorname{res}^{G}_{H}\left(e^{G,A}_{H,\Phi}\right) = \sum_{\substack{\Psi\in\operatorname{Hom}(\operatorname{Hom}(H,A),\mathbb{K}^{\times})\\ [H,\Psi]_{G}=[H,\Phi]_{G}}}e^{H,A}_{H,\Psi}.

2.

ind⁡HGres⁡HG(eH,ΦG,A)=[NG(H):H]eH,ΦG,A.\operatorname{ind}^{G}_{H}\operatorname{res}^{G}_{H}\left(e^{G,A}_{H,\Phi}\right) = [N_G(H):H]e^{G,A}_{H,\Phi}.

3.

ind⁡HG(eH,ΦH,A)=[NG(H,Φ):H]eH,ΦG,A.\operatorname{ind}^{G}_{H}\left(e^{H,A}_{H,\Phi}\right) = [N_G(H,\Phi):H]e^{G,A}_{H,\Phi}.

The conductor of a primitive idempotent

In the following, AA is an abelian group such that tor⁡exp⁡(G)(A)\operatorname{tor}_{\exp(G)}(A) is finite. Let K\mathbb{K} be a field of characteristic zero that is a splitting field for every Hom⁡(H,A)\operatorname{Hom}(H,A), with H≤GH \leq G, and let O\mathcal{O} be its ring of algebraic integers. Consider the primitive idempotent of BQ(G)B_{\mathbb{Q}}(G) corresponding to the GG-th mark,

eGB(G)=∑H∈[SG]aH[G/H],e^{B(G)}_{G} = \sum_{H\in[S_G]}a_H[G/H],

with aH∈Qa_H \in\mathbb{Q}. The following characterization of a primitive idempotent eG,ΦG,Ae^{G,A}_{G,\Phi} is a generalization of that presented by Müller in [7 §3].

Lemma 2.1. Let Φ∈Hom⁡(Hom⁡(G,A),K×)\Phi\in\operatorname{Hom}(\operatorname{Hom}(G,A),\mathbb{K}^{\times}). Then

eG,ΦG,A=1∣Hom⁡(G,A)∣∑λ∈Hom⁡(G,A)Φ(λ−1)xλ.(3)e^{G,A}_{G,\Phi} = \frac{1}{|\operatorname{Hom}(G,A)|} \sum_{\lambda\in\operatorname{Hom}(G,A)} \Phi\left(\lambda^{-1}\right)x_{\lambda}. \tag*{(3)}

where

xλ:=∑H∈[SG]aH[H,λ∣H]G.x_{\lambda}:= \sum_{H\in[S_G]}a_H[H,\lambda|_{H}]_{G}.

Proof. It is straightforward that πK([H,λ∣H]G)=ϕK(G/H)λ∣K\pi_K([H,\lambda|_H]_G)=\phi_K(G/H)\lambda|_K for K≤GK\leq G, hence

πK∘res⁡KG(xλ)=ϕK(eGB(G))λ∣K={λif K=G,0otherwise.\pi_K\circ\operatorname{res}^G_K(x_\lambda)=\phi_K\left(e^{B(G)}_G\right)\lambda|_K= \begin{cases} \lambda& \text{if }K=G,\\ 0 & \text{otherwise.} \end{cases}

For arbitrary (K,Ψ)∈DGA(K,\Psi)\in D^A_G, we have the following: if KK is proper, then

sK,ΨG,A(∑λ∈Hom⁡(G,A)Φ(λ−1)xλ)=∑λ∈Hom⁡(G,A)Φ(λ−1)sK,ΨG,A(xλ)=0;s^{G,A}_{K,\Psi}\left(\sum_{\lambda\in\operatorname{Hom}(G,A)}\Phi(\lambda^{-1})x_\lambda\right)=\sum_{\lambda\in\operatorname{Hom}(G,A)}\Phi(\lambda^{-1})s^{G,A}_{K,\Psi}(x_\lambda)=0;

if K=GK=G, then

sG,ΨG,A(∑λ∈Hom⁡(G,A)Φ(λ−1)xλ)=∑λ∈Hom⁡(G,A)Φ(λ−1)Ψ(λ)={∣Hom⁡(G,A)∣if Ψ=Φ,0otherwise.\begin{aligned} s^{G,A}_{G,\Psi}\left(\sum_{\lambda\in\operatorname{Hom}(G,A)}\Phi(\lambda^{-1})x_\lambda\right)=\sum_{\lambda\in\operatorname{Hom}(G,A)}\Phi(\lambda^{-1})\Psi(\lambda) \\ &=\begin{cases} |\operatorname{Hom}(G,A)| & \text{if }\Psi=\Phi,\\ 0 & \text{otherwise.} \end{cases} \end{aligned}

by row orthogonality relations for the group Hom⁡(G,A)\operatorname{Hom}(G,A), proving that

∣Hom⁡(G,A)∣eG,ΦG,A=∑λ∈Hom⁡(G,A)Φ(λ−1)xλ.|\operatorname{Hom}(G,A)|e^{G,A}_{G,\Phi}=\sum_{\lambda\in\operatorname{Hom}(G,A)}\Phi(\lambda^{-1})x_\lambda.

For a positive integer nn, we set

rn(G):=[G[n]:G′]0gcd⁡(n,[G[n]:G′]0),r^n(G):=\frac{[G^{[n]}:G']_0}{\gcd\left(n,[G^{[n]}:G']_0\right)},

that is, rn(G)r^n(G) is the product of the prime divisors of [G[n]:G′][G^{[n]}:G'] which are coprime to nn. In general, we set

rA(G):=rexp⁡(tor⁡exp⁡(G)(A))(G).r^A(G):=r^{\exp(\operatorname{tor}_{\exp(G)}(A))}(G).

This notation is useful, since the exponent of tor⁡exp⁡(G)(A)\operatorname{tor}_{\exp(G)}(A) may change as GG varies.

Theorem 2.2. Let GG be a finite group, and let AA be an abelian group such that tor⁡exp⁡(G)(A)\operatorname{tor}_{\exp(G)}(A) is finite. Let KK be a field of characteristic zero that is a splitting field for every Hom⁡(H,A)\operatorname{Hom}(H,A), with H≤GH\leq G. For (H,Φ)∈DGA(H,\Phi)\in D^A_G, the conductor of the primitive idempotent eH,ΦG,Ae^{G,A}_{H,\Phi} of BKA(G)B^A_K(G) is

cH,ΦG,A=[NG(H,Φ):H] ∣Hom⁡(H,A)∣ rA(H).(4)c^{G,A}_{H,\Phi}=[N_G(H,\Phi):H]\,|\operatorname{Hom}(H,A)|\,r^A(H). \tag*{(4)}

Proof. Let

tor⁡exp⁡(G)(A)=Cd1×Cd2×⋯×Cdt\operatorname{tor}_{\exp(G)}(A)=C_{d_1}\times C_{d_2}\times\cdots\times C_{d_t}

be a decomposition of tor⁡exp⁡(G)(A)\operatorname{tor}_{\exp(G)}(A) into invariant factors, where di∣di+1d_i\mid d_{i+1} for i=1,…,t−1i=1,\ldots,t-1, and dt=exp⁡(tor⁡exp⁡(G)(A))d_t=\exp\left(\operatorname{tor}_{\exp(G)}(A)\right). We first consider the case H=GH=G, and set

y:=ρ(∣Hom⁡(G,A)∣eG,ΦG,A)=(∑ϕ∈Hom⁡(K,A)aK,ϕϕ)K≤G.y:=\rho\left(|\operatorname{Hom}(G,A)|e^{G,A}_{G,\Phi}\right)=\left(\sum_{\phi\in\operatorname{Hom}(K,A)}a_{K,\phi}\phi\right)_{K\leq G}.

which lies in BOA(G)~\widetilde{B^A_O(G)} by Lemma 2.1. We test Boltje’s integrality criterion on yy. Because the KK-th entry of yy is zero for any proper subgroup KK, we only need to check the congruences for (K,ϕ)(K,\phi) such that NG(K,ϕ)=GN_G(K,\phi)=G and ϕ∈Im⁡(res⁡KG)\phi\in\operatorname{Im}\left(\operatorname{res}^G_K\right), that is, K⊴GK\trianglelefteq G and ϕ=λ∣K\phi=\lambda|_K for some λ∈Hom⁡(G,A)\lambda\in\operatorname{Hom}(G,A). For such a pair, we have

∑(L,ψ)∈MGA(K,ϕ)≤(L,ψ)μ(K,L)aL,ψ=∑λ∈(res⁡KG)−1(ϕ)μ(K,G)Φ(λ−1)={μ(K,G)∣ker⁡(res⁡KG)∣Φ(λ0−1)if ker⁡(res⁡KG)≤ker⁡(Φ),0otherwise,\begin{aligned} \sum_{\substack{(L,\psi)\in\mathcal{M}^A_G\\(K,\phi)\leq(L,\psi)}}\mu(K,L)a_{L,\psi} = \sum_{\lambda\in\left(\operatorname{res}^G_K\right)^{-1}(\phi)}\mu(K,G)\Phi(\lambda^{-1}) \\ &= \begin{cases} \mu(K,G)|\ker(\operatorname{res}^G_K)|\Phi(\lambda_0^{-1}) & \text{if }\ker(\operatorname{res}^G_K)\leq\ker(\Phi),\\ 0 & \text{otherwise}, \end{cases} \end{aligned}

for some λ0∈(res⁡KG)−1(ϕ)\lambda_0\in\left(\operatorname{res}^G_K\right)^{-1}(\phi). Since there is a natural identification Hom⁡(L,A)≅Hom⁡(L,tor⁡exp⁡(G)(A))\operatorname{Hom}(L,A)\cong\operatorname{Hom}\left(L,\operatorname{tor}_{\exp(G)}(A)\right) for L≤GL\leq G, we have

∣ker⁡(res⁡KG)∣=∏i=1t[G:KG[di]]|\ker(\operatorname{res}^G_K)|=\prod_{i=1}^{t}[G:KG^{[d_i]}]

by Lemma 1.2, while Φ(λ0−1)\Phi(\lambda_0^{-1}) is a root of unity. Moreover, every [G:KG[di]][G:KG^{[d_i]}] is a divisor of [G:KG[dt]][G:KG^{[d_t]}]. If [G:KG[dt]][G:KG^{[d_t]}] and [KG[dt]:KG′][KG^{[d_t]}:KG'] are not coprime, then μ(K,G)=0\mu(K,G)=0 by Lemma 1.4. On the other hand, if gcd⁡([G:KG[dt]],[KG[dt]:KG′])=1\gcd([G:KG^{[d_t]}],[KG^{[d_t]}:KG'])=1, every prime divisor of [KG[dt]:KG′][KG^{[d_t]}:KG'] is a divisor of [G[dt]:G′][G^{[d_t]}:G'] that does not divide dtd_t by Lemma 1.1, so it is a divisor of rdt(G)=rA(G)r^{d_t}(G)=r^A(G). Hence, μ(K,G)∣Hom⁡(G,A)∣rA(G)\mu(K,G)|\operatorname{Hom}(G,A)|r^A(G) is multiple of μ(K,G)[G:KG′]\mu(K,G)[G:KG'], and therefore a multiple of [NG(K):K]=[G:K][N_G(K):K]=[G:K], by part 3 of Proposition 1.3. Since NG(K,ϕ)=GN_G(K,\phi)=G, Boltje’s integrality criterion implies that

rA(G)y=ρ(∣Hom⁡(G,A)∣rA(G)eG,ΦG,A)∈ρ(BOA(G)),r^A(G)y=\rho\left(|\operatorname{Hom}(G,A)|r^A(G)e^{G,A}_{G,\Phi}\right)\in\rho\left(B^A_O(G)\right),

hence ∣Hom⁡(G,A)∣rA(G)eG,ΦG,A∈BOA(G)|\operatorname{Hom}(G,A)|r^A(G)e^{G,A}_{G,\Phi}\in B^A_O(G). For general (H,Φ)(H,\Phi), we have that

[NG(H,Φ):H]∣Hom⁡(H,A)∣rA(H)eH,ΦG,A=ind⁡HG(∣Hom⁡(H,A)∣rA(H)eH,ΦH,A)[N_G(H,\Phi):H]|\operatorname{Hom}(H,A)|r^A(H)e^{G,A}_{H,\Phi} =\operatorname{ind}^G_H\left(|\operatorname{Hom}(H,A)|r^A(H)e^{H,A}_{H,\Phi}\right)

belongs to BOA(G)B^A_O(G) since ∣Hom⁡(H,A)∣rA(H)eH,ΦH,A|\operatorname{Hom}(H,A)|r^A(H)e^{H,A}_{H,\Phi} belongs to BOA(H)B^A_O(H). This shows that [NG(H,Φ):H]∣Hom⁡(H,A)∣rA(H)[N_G(H,\Phi):H]|\operatorname{Hom}(H,A)|r^A(H) divides cH,ΦG,Ac^{G,A}_{H,\Phi}. Conversely, note that tor⁡exp⁡(H)(A)\operatorname{tor}_{\exp(H)}(A) is also finite, and consider a decomposition

tor⁡exp⁡(H)(A)=Cf1×Cf2×⋯×Cfs\operatorname{tor}_{\exp(H)}(A)=C_{f_1}\times C_{f_2}\times\cdots\times C_{f_s}

into invariant factors, where fi∣fi+1f_i \mid f_{i+1} for i=1,…,s−1i=1,\ldots,s-1 and fs=exp⁡(tor⁡exp⁡(H)(A))f_s=\exp(\operatorname{tor}_{\exp(H)}(A)). Note that the coefficient of [H,1]G[H,1]_G in eH,ΦG,Ae^{G,A}_{H,\Phi} with respect to the monomial basis is ([NG(H,Φ):H]∣Hom⁡(H,A)∣)−1\left([N_G(H,\Phi):H]|\operatorname{Hom}(H,A)|\right)^{-1}, therefore [NG(H,Φ):H]∣Hom⁡(H,A)∣[N_G(H,\Phi):H]|\operatorname{Hom}(H,A)| divides cH,ΦG,Ac^{G,A}_{H,\Phi}. We now prove that rA(H)=rfs(H)r^A(H)=r^{f_s}(H) divides

u(H,Φ):=cH,ΦG,A[NG(H,Φ):H]∣Hom⁡(H,A)∣.u(H,\Phi):=\frac{c^{G,A}_{H,\Phi}}{[N_G(H,\Phi):H]|\operatorname{Hom}(H,A)|}.

Let pp be a prime divisor of [H[fs]:H′][H^{[f_s]}:H'] that does not divide fsf_s. For any K≤HK\leq H such that [H:K]=p[H:K]=p, we have [H:KH[fs]]=1[H:KH^{[f_s]}]=1, hence [H:KH[fi]]=1[H:KH^{[f_i]}]=1 for each ii, and so

∣ker⁡(res⁡KH)∣=∏i=1s[H:KH[fi]]=1,\left|\ker\left(\operatorname{res}^{H}_{K}\right)\right|=\prod_{i=1}^{s}[H:KH^{[f_i]}]=1,

proving that res⁡KH:Hom⁡(H,A)→Hom⁡(K,A)\operatorname{res}^{H}_{K}:\operatorname{Hom}(H,A)\to\operatorname{Hom}(K,A) is injective. Let

Ip={K≤H∣[H:K]=p}.\mathcal{I}_p=\{K\leq H\mid[H:K]=p\}.

We have ∣Ip∣≡1(modp)|\mathcal{I}_p|\equiv1\pmod p, and Ip\mathcal{I}_p splits into equivalence classes under the relation =G=_{G}. Therefore, there is some K∈IpK\in\mathcal{I}_p such that pp does not divide

m=#{L∈Ip∣L=GK}.m=\#\{L\in\mathcal{I}_p\mid L=_{G}K\}.

For such KK, the coefficient of [K,1]G[K,1]_G in [NG(H,Φ):H]∣Hom⁡(H,A)∣eH,ΦG,A[N_G(H,\Phi):H]|\operatorname{Hom}(H,A)|e^{G,A}_{H,\Phi} with respect to the monomial basis is

1∣H∣∑L≤H,L=GK,ϕ∈ker⁡(res⁡LH)∣L∣μ(L,H)Φ(ϕ−1)=−∣K∣∣H∣∑L≤H,L=GKΦ(1)=−mp.\frac{1}{|H|}\sum_{\substack{L\leq H,\\L=_{G}K,\\\phi\in\ker(\operatorname{res}^{H}_{L})}}|L|\mu(L,H)\Phi(\phi^{-1})=-\frac{|K|}{|H|}\sum_{\substack{L\leq H,\\L=_{G}K}}\Phi(1)=-\frac{m}{p}.

Since u(H,Φ)(−mp)u(H,\Phi)\left(-\frac{m}{p}\right) must be an integer, we conclude that pp must divide u(H,Φ)u(H,\Phi), completing the proof. □

Corollary 2.3. Let GG be a finite group, and let nn be a positive integer. For (H,hH[n])∈DGCn(H,hH^{[n]})\in D_G^{C_n}, the conductor of eH,hG,ne^{G,n}_{H,h} is

cH,hG,n=[NG(H,hH[n]):H[n]] rn(H).(5)c^{G,n}_{H,h}=[N_G(H,hH^{[n]}):H^{[n]}]\,r^n(H). \tag*{(5)}

For both the Burnside ring and the monomial representation ring, the order of a group can be easily recovered from their conductors [7 Proposition 3.7]. This also holds for all fibered Burnside rings with cyclic fiber group.

Proposition 2.4. Let nn be a positive integer. For a finite group GG, we have

∣G∣=max⁡{cH,hG,n∣(H,hH[n])∈DGCn}=min⁡{t∈Z>0∣teH,hG,n∈BOA(G), ∀ (H,hH[n])∈DGCn}.|G|=\max\left\{c^{G,n}_{H,h}\mid(H,hH^{[n]})\in D_G^{C_n}\right\} =\min\left\{t\in\mathbb{Z}_{>0}\mid te^{G,n}_{H,h}\in B^A_{\mathit O}(G),\ \forall\,(H,hH^{[n]})\in D_G^{C_n}\right\}.

Proof. Take a characteristic zero field K\mathrm{K} containing a primitive nn-th root of unity. The conductors of the primitive idempotents of BK(G)B_{\mathrm{K}}(G) do not depend on the field K\mathrm{K}, and [NG(H,hH[n]):H[n]]rn(H)[N_G(H,hH^{[n]}):H^{[n]}]r^n(H) divides ∣G∣|G| for every pair (H,hH[n])(H,hH^{[n]}). As cH,hG,n=∣G∣c^{G,n}_{H,h}=|G| is reached at least once, this proves both equalities. □

The previous argument fails for more general fibers, as shown in the following example.

Example 2.5. Let G=C4G=C_4, H≤C4H\leq C_4 be the subgroup of order 2 and A=C2×C2A=C_2\times C_2. Then NG(H,1)=NG(H)=C4N_G(H,1)=N_G(H)=C_4, ∣Hom⁡(H,C2×C2)∣=4|\operatorname{Hom}(H,C_2\times C_2)|=4, and rC2×C2(H)=1r^{C_2\times C_2}(H)=1. Thus, we have

cH,1C4,C2×C2=[C4:H]∣Hom⁡(H,C2×C2)∣rC2×C2(H)=8,c^{C_4,C_2\times C_2}_{H,1}=[C_4:H]|\operatorname{Hom}(H,C_2\times C_2)|r^{C_2\times C_2}(H)=8,

which exceeds $|C_4|=4.

The idempotents of B(G)B(G) whose conductors are exactly ∣G∣|G| play a central role in the proof of the existence of normalizing isomorphisms, with important consequences on the isomorphism problem of the Burnside ring [9]. In the case of D(G)D(G), we see that GG is abelian if and only if all conductors are equal to ∣G∣|G| [7]. More generally, we can characterize pairs (H,hH[n])(H,hH^{[n]}) for which cH,hG,n=∣G∣c^{G,n}_{H,h}=|G|.

Proposition 2.6. Let nn be a positive integer. For (H,hH[n])∈DGCn(H,hH^{[n]})\in D_G^{C_n}, cH,hG,n=∣G∣c^{G,n}_{H,h}=|G| if and only if HH is a normal abelian subgroup such that H[n]H^{[n]} has square-free order coprime to nn and hH[n]∈Z(G/H[n])hH^{[n]}\in Z(G/H^{[n]}).

Proof. Note that [NG(H,hH[n]):H[n]]rn(H)=[G:1][N_G(H,hH^{[n]}):H^{[n]}]r^n(H)=[G:1] implies NG(H,hH[n])=GN_G(H,hH^{[n]})=G, and so NG(H)=GN_G(H)=G and hH[n]∈Z(G/H[n])hH^{[n]}\in Z(G/H^{[n]}). Then [G:H[n]]rn(H)=[G:1][G:H^{[n]}]r^n(H)=[G:1] implies [H[n]:1]=rn(H)[H^{[n]}:1]=r^n(H), the latter is a divisor of [H[n]:H′][H^{[n]}:H'], thus implying H′=1H'=1 and proving the assertion about the order of H[n]H^{[n]}. Conversely, suppose that HH is normal and abelian, and that H[n]H^{[n]} has square-free order coprime to nn. For hH[n]∈Z(G/H[n])hH^{[n]}\in Z(G/H^{[n]}), we have NG(H,hH[n])=GN_G(H,hH^{[n]})=G, and, since H′=1H'=1, we have rn(H)=∣H[n]∣r^n(H)=|H^{[n]}|. Consequently, cH,hG,n=[G:H[n]]∣H[n]∣=∣G∣c^{G,n}_{H,h}=[G:H^{[n]}]|H^{[n]}|=|G|. □

Specializing to pp-groups, we get the following.

Corollary 2.7. For a prime number pp, let PP be a pp-group and let qq be a power of pp. For (H,hH[q])∈DPCq(H,hH^{[q]})\in D_P^{C_q}, cH,hP,q=∣P∣c^{P,q}_{H,h}=|P| if and only if HH is a normal abelian subgroup such that H[q]=1H^{[q]}=1 and h∈Z(P)h\in Z(P).

Proof. This follows immediately from the previous result, since in this case $r^q(H)=1. □

References

  1. [1]R. Boltje, Integrality conditions for elements in ghost rings of generalized Burnside rings. Preprint, 2004. https://boltje.math.ucsc.edu/publications/p04z.pdf
  2. [2]R. Boltje, D. Yılmaz, The A-fibered Burnside ring as A-fibered biset functor in characteristic zero. Algebras and Representation Theory, 24(5):1359–1385, 2021.arxiv.org/abs/1909.06668
  3. [3]A. W. M. Dress, E. Vallejo, A simple proof for a result by Kratzer and Thévenaz concerning the embedding of the Burnside ring into its ghost ring. Journal of Algebra, 154(2):356–359, 1993.DOI
  4. [4]B. García, A. G. Raggi-Cárdenas, On fibered Burnside rings, fiber change maps and cyclic fiber groups. Journal of Pure and Applied Algebra, 229(6), article 107961, 2025. https://doi.org/10.1016/j.jpaa.2025.107961
  5. [5]T. Hawkes, I. M. Isaacs, M. Özaydin, On the Möbius function of a finite group. Rocky Mountain Journal of Mathematics, 19(4):1003–1034, 1989. https://doi.org/10.1216/RMJ-1989-19-4-1003
  6. [6]C. Kratzer, J. Thévenaz, Fonction de Möbius d’un groupe fini et anneau de Burnside. Commentarii Mathematici Helvetici, 59:425–438, 1984. https://doi.org/10.5169/seals-45403
  7. [7]M. Müller, On the isomorphism problem for the ring of monomial representations of a finite group. Journal of Algebra, 333(1):427–457, 2011. https://doi.org/10.1016/j.jalgebra.2011.02.013
  8. [8]D. M. Nicolson, The orbit of the regular G-set under the full automorphism group of the Burnside ring of a finite group G. Journal of Algebra, 51:288–299, 1978.DOI
  9. [9]A. G. Raggi-Cárdenas, L. Valero-Elizondo, Normalizing isomorphisms between Burnside rings. Journal of Algebra, 277(2):643–657, 2004. https://doi.org/10.1016/j.jalgebra.2003.10.031

Paper details

vlnc:W_2uerwfsf
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 3

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents