2
votes

I've made an iOS App to range the beacons. I've noticed that the RSSI value from beacons is fluctuating randomly with time. In order to get smooth RSSI value, I am trying to use Kalman filter.

In Kalman filter equations as described here, the measurement noise (R) can be calculated by measuring variance from series of RSSI values and the process noise (Q) can be assumed as negligible. However, I couldn't figure out exact idea about the estimate of error variance (P) in the equation.

Since my actual measurement data is the series of RSSI values, how am I supposed to implement Kalman filter?

1

1 Answers

1
votes

Basically the estimate of error variance(P) depends on it's own past value and the process noise(Q). Since the process noise(Q) is negligible or very small value (0.00001) and P doesn't depend on actual measurement, eventually it becomes a fixed value. Moreover if your system is mobile, you can take series of RSSI values and your movement as input variables.