Sorry, I don't really know what this could be caused by. I would start by looking at the data I get from the system at various points and compare it with measurements from the physical world to see if it seems reasonable. I would also input simulated TDoA data into the estimator to see how it behaves, but it sounds like you already have made simulations.kind of like a scale model. So getting away the error gets bigger.
Maybe double check the anchor positions? (You probably already did )
As far as I know I works fairly well all the way to the edge. We use data from the accelerometers as well and maybe that helps a bit if the LPS data is less good?Also, how much near the side of the cube can you go getting good measures? To me they start to get bad already about a meter from it.
Not sure @Arnaud is back from his vacation, maybe he can help us here?why do you use the light' speed in the void? Shouldn't it be the light' speed in the air
No, they look good to me. We have had much smaller set ups than this, down to around 2.5x2.5x2.0 meters. I think there are some non linearities when you get close to the anchors but from my experience, your space should be OK.Do you think my anchors are too close?