The Andromeda galaxy is observed to have a system of two large dwarf ellipticals and \sim 13 smaller satellite galaxies that are currently co-rotating in a thin plane , in addition to 2 counter-rotating satellite galaxies .
We explored the consistency of those observations with a scenario where the majority of the co-rotating satellite galaxies originated from a subhalo group , where NGC 205 was the host and the satellite galaxies occupied dark matter sub-subhalos .
We ran N-body simulations of a close encounter between NGC 205 and M31 .
In the simulations , NGC 205 was surrounded by massless particles to statistically sample the distribution of the sub-subhalos expected in a subhalo that has a mass similar to NGC 205 .
We made Monte Carlo samplings and found that , using a set of reference parameters , the probability of producing a thinner distribution of sub-subhalos than the observed NGC 205 + 15 smaller satellites ( thus including the 2 counter-rotators , but excluding M32 ) increased from < 10 ^ { -8 } for the initial distribution to \sim 10 ^ { -2 } at pericentre .
The probability of the simulated sub-subhalos occupying the locations of the observed co-rotating satellites in the line of sight velocity versus projected on-sky distance plane is at most 2 \times 10 ^ { -3 } for 11 out of 13 satellites .
Increasing the mass of M31 and the extent of the initial distribution of sub-subhalos gives a maximum probability of 4 \times 10 ^ { -3 } for all 13 co-rotating satellites , but the probability of producing the thinness would drop to \sim 10 ^ { -3 } .