Fellow RFID forum users and myself are looking for some help deriving position from RSSI values. The basics are clear to me, but I really want to figure out a rock solid formula or algorithm that will allow for any number of readers, as long as their position is known and properly weighted in the equation.
Actually, I’m imagining two types of scenarios, but the first one is what I want to shoot for at the moment. This first scenario is an area or room with RFID readers in known, fixed positions. This means that the tags will move around within the controlled area, and positions will be calculated in terms of X,Y. The method of calculating these coordinates must be flexibile enough that a random number of receivers (3+) can be randomly distributed across an area, and be able to handle situations where some receivers report no RSSI values because tags are out of range.
The second, much more ambitious task is to come up with a calculation that can handle a much more free form mesh network of RSSI values to derive X,Y, and possibly even Z coordinates of each tag. I imagine this to be a solution where we have an open free form space (think open desert) where each tag would also be a reader, and would report back RSSI values for all its neighbors. There would probably need to be a few fixed readers throughout the space, but not many… not enough to cover the whole space at least. That means that relative positions would have to be calculated as they pertain to the fixed readers, and to each other.
Anyone interested in starting a project based on these goals?