Rssi kalman filtering for trilateration?

Archive
1

hello,
i am trying to calculate the position of android phones by measuring the rssi signal strength of several beacons with ~2 meters distance to each other.
the measured signals are quite noisy and unpredictable and sometimes the connection drops so i try to smooth them with an lowpass filter.
but still the measurement is not good for a simple trilateration or something with more than three beacons.
i guess implementing a one dimensional kalman filter would be right method for getting rid of the errors in measurement ?
are there any plans to include such smoothing and tracking functions into the api or is there already any smoothing on the measurements ?
do have any experience with rssi smoothing (kalman) and trilateration you could share ?
what could be the ideal scan period for this ?

thanks a lot

2

We do have experience with that. More over we are experimenting with that and please do expect kalman filtering within next releases of the SDK.

3

how do you configure the kalman filter ? is it one dimensional or is some sensor fusion involved ? in my experience the kalman on only one value (accuracy) behaves like an simple lowpass filter, so i switched to that. but still not realy happy with the results.

4

I'll refrain myself from disclosing any other information right now. I hope that you understand that.