The speed of light is determined from round trip measurements: measuring a one way trip hits circularity issues in establishing time. So, bounce a pulse of light off a mirror at unit distance (i.e. 1 light second) and it takes two seconds to return but we don't know that $t$ on the way out is the same as $t'$ on the way back.
I wondered what happens with a triangle of points at unit distance. Now we have six unknowns ($t, t'$ along each edge) and three there and back circuits as well as one circuit round all three points, but in two directions: altogether five equations in six unknowns - not quite enough.
How about a tetrahedron ? Six edges, so twelve unknowns, 6 round trip equations on the edges, and taking three of the four faces a further six equations for the triangular round trips (in both directions). Twelve equations in twelve unknowns with the obvious solution $t_1 = t_1'$ etc.
I was about to claim my Nobel physics prize when a thought occurred to me.
The paths representing the twelve circuits can be expressed as $1's$ and $0's$ in a $12 \times 12$ matrix $M$, and the transit times in the form $(t_1, t_1',...)M = (2,2,2,2,2,2,3,3,3,3,3,3)$. In order to ensure that $t_1 = t_1'$ etc is the unique solution we need $M$ to be non-singular.
Bad news, $Det(M) = 0$ (thanks to https://matrix.reshish.com/determinant.php).
All was not yet lost as the tetrahedron has another face with another two equations, so fourteen equations in twelve unknowns represented by $M'$ a $14 \times 12$ matrix. But $Rank(M') = 9$ (thanks again https://matrix.reshish.com/rankCalculation.php), i.e only nine of the equations are linearly independent. Cancel Nobel prize for time being.
It is generally accepted that the one way speed of light cannot be determined, and this seems to infer the following graph theory proposition .....
"In any directed graph (where one way transit times can depend on direction), knowing the transit times around all closed circuits (in both directions) is insufficient to determine any single one way transit time"
So is this true, is it proven ?