Periodically nonuniform sampling (PNS) is employed to sample an interesting component over a subband in a bandlimited signal exclusive of using a physical filter. By controlling the sampling clock and the interpolation algorithm with respected to the interesting component, the sampling and frequency selection can be implemented simultaneously. The sampling rate and order are considered to merely allow the successful reconstruction of the interesting component and the average sampling rate is desirable to be lower than that to sample the entire signal. By analyzing the possible spectral aliasing cases in the selected band, it is found that the number of spectral replicas is at most 2 less than the maximum number in any subband of the signal. In addition, through the analysis, it is found that whatever the support of the selected band is, the average sampling rate has a lower bound. The limitation is primarily determined by the bandwidth of the complete signal.
[1]
D. Linden.
A Discussion of Sampling Theorems
,
1959,
Proceedings of the IRE.
[2]
H. Landau.
Necessary density conditions for sampling and interpolation of certain entire functions
,
1967
.
[3]
Gennady Y. Poberezhskiy,et al.
Signal reconstruction technique allowing exclusion of antialiasing filter
,
2001
.
[4]
A. Kohlenberg.
Exact Interpolation of Band‐Limited Functions
,
1953
.
[5]
Alan J. Coulson,et al.
A generalization of nonuniform bandpass sampling
,
1995,
IEEE Trans. Signal Process..
[6]
Y. Poberezhskiy,et al.
Sampling technique allowing exclusion of antialiasing filter
,
2000
.