Digital Alias-free Signal Processing

General points
DASP stands for Digital Alias-free Signal Processing.

DASP represents advanced DSP techniques for digital processing of signals in a considerably extended frequency range not necessarily limited by the used sampling rate as usual.
It should be emphasized that the remarkable features of DASP are basically due to paying special attention to the original signal digitization operations and to the usage of deliberate pseudorandomization as a tool for achieving specific positive effects. DASP, specifically:

Is recognizing the significance of the sampling and quantization processes typically underestimated by the traditional DSP techniques
Makes the digitizing operations substantially more flexible and enrich them with additional features by using deliberate pseudorandomization procedures when appropriate
Adapts the sampling and quantization operations to the specifics of the particular signal processing case
Performs sampling either periodically or nonuniformly if and when aliasing has to be avoided
Performs quantization either as a deterministic or pseudorandomized operation depending on the requirements of the given specific signal processing case
Is using special signal processing algorithms matched to the specifics of the digitizing mode chosen for the particular signal processing case including algorithms suitable for processing non-equispaced signal sample value sequences

So the DASP system design philosophy is based on optimization of signal digitizing made considerably more flexible by applying the deliberately pseudo-randomization approach to both sampling and/or quantization operations. Most often, special algorithms are needed for processing signals digitized in this way. While the most spectacular effect of alias suppression is achieved by pseudo-randomization of sampling, pseudo-randomized quantizing often is quite useful too, and this additional facet can be brought to bear even when sample timings are uniform.
DASP capabilities, advantages and limitations are demonstrated by the DASP Lab System, Winner of the European Information Technology Prize`97.

DASP- Lab System, serving well as a Demonstrator for DASP, is a family of PC-based hardware/software Virtual Instruments for fully digital analysis of RF signals in Time, Frequency and Modulation Domains in the frequency range from dc up to 1.2 GHz. Operations in so wide a frequency range are supported by the usage of nonuniform sampling. And the point is that the highest mean sampling rate for this system is only 80 MS/sec. Therefore the fully digital alias-free signal analysis in this case is carried out in a frequency range 30 times exceeding the bandwidth achievable under the same conditions by the classic DSP.

While this system certainly helps to reveal the DASP capabilities, there is more worth knowing about DASP than might be seen from working with DASP Lab System. It should be realised, for instance, that it does not demonstrate the DASP real-time applications. And they probably are the most promising and important.

Thus DASP possess significant beneficial advantages and it has a strong application potential.

Apparently DASP also has its limitations. However they substantially differ from the limitations of traditional DSP.

The DASP origins can be traced back to 1960`ies. Many outstanding researchers have contributed to the theoretical fundaments of this technology. The theoretical fundaments of DASP to some extent are given in the book: Randomized Signal Processing by I. Bilinskis, A. Mikelsons, Prentice Hall International.

DASP does not contradict the classical DSP. It rather supplements DSP as it is generally known. Actually theory and techniques of modern DSP should be reconsidered with the knowledge accumulated in the area of DASP blended with the classical concepts of DSP in order to gain a comprehensive picture.