Group intersection PTIME equals set-stabilizer

From Groupprops

This article gives the statement, and possibly proof, of a polynomial-time equivalence between two computational/decision problems in computational group theory (or related areas such as graph theory

Statement

The group intersection problem for subgroups of the permutation group is polynomial-time equivalent to the set stabilizer problem.

Definitions of the problems

Group intersection problem

Here, two subgroups G1 and G2 of the symmetric group on a set S are specified by means of generating sets A1 and A2, and we need to find a small generating set for G1 ∩ G2.

Set-stabilizer problem

Here, we have a set S and a group G acting on S. Let T be a subset of S. We need to find the subgroup of G comprising those elements which stabilize T as a set.

Proof

Set stabilizer reduces to group intersection

Consider the data (G,S,T) as given for the set stabilizer problem. Now, send the following data to the group intersection problem:

  • Set S to be the same as the S for the set stabilizer problem
  • Set G1=G
  • Set G2=Sym(T) × Sym(Tc) viz G2 is the stabilizer of S in the whole of the symmetric group

Note that G2 is actually a group with a very small generating set (a generating set of size 4).

The output given by this problem is precisely the same as the set stabilizer of S in G.

Note that this is a many-one reduction.

Group intersection reduces to set stabilizer

Consider the data (G1,G2,S) as given for the group intersection problem. Consider the set S2 with G1 acting on the first coordinate and G2 acting on the second coordinate.

Then send the following data to the set stabilizer problem:

  • Set S of the set stabilizer problem to S2
  • Set G of the set stabilizer problem to be G1 × G2
  • Set T of the set stabilizer problem as the set of diagonal elements in S2

The set stabilizer of T is thus precisely those pairs of elements (g,h) such that for every s in S, g.s=h.s. This forces the set stabilizer of T to comprise only elements of the form (g,h) where g=h, which forces the set stabilizer to be the diagonal subgroup of G1 ∩ G2 × G1 × G2. Finding a generating set for this diagonal subgroup is the same as finding a generating set for G1 ∩ G2.