One-Dimensional Missing-Data APES via Expectation Maximization - Spectral Analysis of Signals: The Missing Data Case - page 54

4

True Data
Interpolated Data
Missing Data Locations

2

0

2

−

− 4

20 40 60 80 100 120

n

(a)

4

True Data
Interpolated Data
Missing Data Locations

2

0

− 2

− 4

20 40 60 80 100 120

n

(b)

4

True Data
Interpolated Data
Missing Data Locations

2

0

− 2

− 4

20 40 60 80 100 120

n

(c)

4

True Data
Interpolated Data
Missing Data Locations

2

0

− 2

− 4

20 40 60 80 100 120

n

(d)

= ,σ

n

= .

FIGURE 5.3 : Interpolation of the missing samples [ N 01, 51 (40%) miss-

ing samples]. (a) Real part of the data interpolated via MAPES-EM1, (b) imaginary part of
the data interpolated via MAPES-EM1, (c) real part of the data interpolated via MAPES-
EM2, and (d) imaginary part of the data interpolated via MAPES-EM2.

128 0