I am working on an accelerometer on a mobile device. When an event occurs, I record around 30 acceleration values. The accelerometer, being extremely noisy is very unreliable. What makes it worse is that my motion during the acquisition time is also of the same order as the noise. Is there a good method to estimate the curve the mobile goes through with a decent accuracy?
Edit: I am trying to estimate the position of the mobile. Further, the amplitudes of acceleration are of the same order as the noise(I logged the accelerometer values when the mobile was stationary and made this estimate). Further, The noise is approximately gaussian.