Digital Alias-free Signal Processing

10

Hybrid Sampling

Both uniform periodic and nonuniform randomized sampling have their advantages and disadvantages. They are specific and different. However, it can be said that the drawbacks of periodic sampling are countered to some extent by advantages of nonuniform randomized sampling and vice versa. Recognition of this situation has led to attempts to synthesize sampling methods that have integrated positive features of uniform and nonuniform sampling methods and at the same time would not be impaired by their drawbacks. This kind of periodic and pseudo-randomized nonuniform sampling hybrids is discussed in this chapter.

Although purposeful development of hybrid sampling is an activity that began only recently, the basic principles of this approach have been in use for a long time. For instance, pseudo-randomized sampling may be mentioned as a case where a periodic clocking process is combined with pseudo-random skipping of the sampling pulses. Therefore it is not very easy to draw a line between pseudorandomized and hybrid sampling models. In fact they partly overlap. The hybrid sampling techniques discussed here are more elaborate and specific than the basic pseudo-randomized sampling techniques.

These sampling techniques also offer an additional opportunity to improve the certainty that the signal sample values will be taken at exact predetermined time instants. This is essential for high-performance signal processing as the sampling instants then have to be known with ...