Usually, signals in acquisition are considered bandlimited and are processed according to the Nyquist condition. However, subsequent sampling values are highly redundant for various signal types, often leading to an additional digital compression step after sampling (e.g., for image data: RAW to JPEG, for audio data: WAV to MPEG).

The theory of compressed sensing describes the possibility to achieve data acquisition with sampling rates far below the Nyquist rate – without any loss of information. For that, existent prior knowledge is employed in the measuring principle as well as in subsequent reconstruction of relevant information. This requires a complete innovative approach to signal acquisition. The eventual aim is to extract a preferably large amount of for the measurement relevant information with every sampling value of the incoming signal. Every measured value contains cumulated information on the holistic signal. The subsequent reconstruction process aims at reconstructing this relevant information of the acquired measured values exclusively.

Ultimately, the objective is to reduce the gathered data through adaptation of the signal acquisition and an increased effort in reconstruction.