Stratifying modular representations of finite groups. (Q661921): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
Created claim: Wikidata QID (P12): Q105419626, #quickstatements; #temporary_batch_1712446058616
Property / Wikidata QID
 
Property / Wikidata QID: Q105419626 / rank
 
Normal rank

Revision as of 00:30, 7 April 2024

scientific article
Language Label Description Also known as
English
Stratifying modular representations of finite groups.
scientific article

    Statements

    Stratifying modular representations of finite groups. (English)
    0 references
    0 references
    0 references
    0 references
    11 February 2012
    0 references
    Let \(G\) be a finite group and let \(k\) be a field of characteristic \(p\). Denote by \(\mathsf{Mod}(kG)\) the category of all \(kG\)-modules. The cohomology ring \(H^*(G,k)\) is graded commutative, and therefore it is possible to form the projective space \(\text{Proj}(H^*(G,k))\) of all non trivial homogeneous prime ideals on this ring. The paper under review shows that there is a bijection between the subsets of \(\text{Proj}(H^*(G,k))\) and those non zero subcategories of \(\mathsf{Mod}(kG)\) which are invariant under direct summands, direct sums, tensoring with simple modules, and satisfy that whenever two objects in an exact sequence are in the category, then is the third. The proof is highly involved and passes through various intermediate steps which are always interesting in their own right. The bijection is defined in the following way. Let \(R\) be a graded commutative ring. Then, under certain assumptions, in any triangulated \(R\)-category \(T\) we get a localisation functor \(L_V\) so that \(L_V\) vanishes precisely on those \(X\) with \(R_q\otimes_R\Hom_T(-,X[n])=0\) for all \(n\) and all primes \(q\) not in \(V\). It can be shown that there is a functor \(\Gamma_V\) and an exact triangle \(\Gamma_VX\rightarrow X\rightarrow L_V\rightarrow\). For a graded commutative ring \(R\) and a prime ideal \(p\), let \(V\) and \(W\) be two sets of primes, both closed under taking primes bigger than any given in the set, so that \(V\setminus W\) has only the element \(p\). Then define \(\Gamma_p:=\Gamma_VL_W\). Given a localising subcategory \(C\) of \(T\), then define \(\sigma(T)\) the primes \(p\) of \(R\) so that \(\Gamma_pC\neq 0\). The map \(\sigma\) gives the bijection, for various \(R\) and \(T\) adapted to the situation in the proof. The inverse is given explicitly in a similar fashion. In the first step \(T\) is the derived category of a graded polynomial algebra \(R\) in \(r\) variables. The second step uses as \(T\) the homotopy category of injective modules over the exterior algebra with \(r\) variables, and \(R\) the Ext-algebra of the trivial module. Then in further steps one passes to the homotopy category of injective modules over a maximal elementary Abelian \(p\)-subgroup of \(G\), and further to homotopy category of injective modules over \(kG\), and finally to the module category of \(kG\). The bijection is always given by some \(\sigma\), adapted to each case.
    0 references
    modular representation theory
    0 references
    group cohomology rings
    0 references
    thick localising subcategories
    0 references
    derived categories
    0 references
    exterior algebra
    0 references
    finite groups
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references