# Non-“weakly group theoretical” integral fusion categories?

Can you exclude integral fusion categories of global dimension 210, such that the simple objects have dimensions {1,5,5,5,6,7,7} and the following fusion matrices (I don't write the trivial one) ?

$\begin{pmatrix} 0 & 1 & 0 & 0& 0& 0& 0 \newline 1 & 1 & 0 & 1& 0& 1& 1 \newline 0 & 0 & 1 & 0& 1& 1& 1 \newline 0 & 1 & 0 & 0& 1& 1& 1 \newline 0 & 0 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 1 & 0& 0& 0& 0 \newline 0 & 0 & 1 & 0& 1& 1& 1 \newline 1 & 1 & 1 & 0& 0& 1& 1 \newline 0 & 0 & 0 & 1& 1& 1& 1 \newline 0 & 1 & 0 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 1& 0& 0& 0 \newline 0 & 1 & 0 & 0& 1& 1& 1 \newline 0 & 0 & 0 & 1& 1& 1& 1 \newline 1 & 0 & 1 & 1& 0& 1& 1 \newline 0 & 1 & 1 & 0& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 0& 1& 0& 0 \newline 0 & 0 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 0 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 0& 1& 1& 1 \newline 1 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 2& 1 \newline 0 & 1 & 1 & 1& 1& 1& 2 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 0& 0& 1& 0 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 2& 1 \newline 1 & 1 & 1 & 1& 2& 0& 3 \newline 0 & 1 & 1 & 1& 1& 3& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 0& 0& 0& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 2 \newline 0 & 1 & 1 & 1& 1& 3& 1 \newline 1 & 1 & 1 & 1& 2& 1& 2 \end{pmatrix}$

or also (just a little variation for the 7-dimensional simple objects):

$\begin{pmatrix} 0 & 1 & 0 & 0& 0& 0& 0 \newline 1 & 1 & 0 & 1& 0& 1& 1 \newline 0 & 0 & 1 & 0& 1& 1& 1 \newline 0 & 1 & 0 & 0& 1& 1& 1 \newline 0 & 0 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 1 & 0& 0& 0& 0 \newline 0 & 0 & 1 & 0& 1& 1& 1 \newline 1 & 1 & 1 & 0& 0& 1& 1 \newline 0 & 0 & 0 & 1& 1& 1& 1 \newline 0 & 1 & 0 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 1& 0& 0& 0 \newline 0 & 1 & 0 & 0& 1& 1& 1 \newline 0 & 0 & 0 & 1& 1& 1& 1 \newline 1 & 0 & 1 & 1& 0& 1& 1 \newline 0 & 1 & 1 & 0& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 0& 1& 0& 0 \newline 0 & 0 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 0 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 0& 1& 1& 1 \newline 1 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 2& 1 \newline 0 & 1 & 1 & 1& 1& 1& 2 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 0& 0& 1& 0 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 2& 1 \newline 1 & 1 & 1 & 1& 2& 1& 2 \newline 0 & 1 & 1 & 1& 1& 2& 2 \end{pmatrix}$ $\begin{pmatrix} 0 & 0 & 0 & 0& 0& 0& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 1 \newline 0 & 1 & 1 & 1& 1& 1& 2 \newline 0 & 1 & 1 & 1& 1& 2& 2 \newline 1 & 1 & 1 & 1& 2& 2& 1 \end{pmatrix}$

Note that 210 = 2.3.5.7 and that these matrices are self-dual and irreducibles. They also commute.
By arXiv:0809.3031 proposition 9.11, if such integral fusion categories exist, they couldn't be "weakly group theoretical", and by arXiv:1208.0840 corollary 6.16, they would be abelian but not braided.
Thank you to Eric Rowell and Leonid Vainermann for these references.
Also thank you to Dave Penneys for asking Eric.
I'm actually looking for a solution of their pentagonal equation, so if anyone can exclude these categories...

Edit about the original motivation :
These matrices are naturally came from my will of classifying the cyclic subfactors:
The first case I consider is "depth 2, irreducible, finite index", i.e. finite dimensional C*-Hopf algebras (also called Kac algebras). The first question to answer is:
Are there non-trivial cyclic Kac algebras ? If so, the first example is certainly maximal.
Now a Kac algebra gives a unitary integral fusion category, so I have written an algorithm investigating all the integral fusion rings of a restrictive class containing necessarily those related to the non-trivial maximal Kac algebras.
For now, I discovered eight new such fusion rings in dimensions 210, 360 and 660.
There are finitely many possibilities for each dimension, but the program is quite long, so that I'm actually working on the optimization of the algorithm.

-
I am interested to know how you got those matrices: where do they come from? –  André Henriques Jul 4 '13 at 21:16
For answering your question, I have edited a final part. –  Sébastien Palcoux Jul 5 '13 at 8:53
Kac algebras are not completely given by a unitary fusion category. You can have two non isomorphic Kac algebras with the same tensor category of representation, for example, a Drinfeld twist deformation of a group algebra. –  César Galindo Jul 12 '13 at 16:18
@CésarGalindo : Do you confirm they have the same fusion category (not only the same fusion ring). Maybe the fusion categories of the Kac algebra and its dual, give completely the Kac algebras. It's true for the few twist deformation of group algebras I know, but I don't know if it's true in general, and you ? –  Sébastien Palcoux Sep 10 '13 at 14:58
Yes, twist deformations of finite dimensional Hopf algebras have the same category of representation, look for example arxiv.org/pdf/math/0107167.pdf A finite dimensional Hopf algebra is completely determined by their category of representation and its fiber functor. –  César Galindo Sep 12 '13 at 22:13