Algorithm 2$f_{WT}\left(\right)$: Measurement noise variance by wavelet transform.

Input:       $S_{sensor}$: a numeric sequence of sensor data, ($n=|S_{sensor}|$);       f: the number of wavelet coefficients to be selected; Output:       R: measurement noise variance of $S_{sensor}$; begin       (1) $S^{wt}:=$ WaveletTransform$\left(S_{sensor}\right)$;       (2) $S^{wt}:=$ Concatenate$(S^{wt}[1:f],$Array.zeros$(n-f))$;     // Take high energy coefficients.       (3) $S_{estimate}:=$ InverseWaveletTransform$\left(S^{wt}\right)$;       (4) $R:=$ CalculateVAR$(S_{sensor},S_{estimate})$;     // Calculate the variance between two sequences.       (5) return R; end