Angličtina
This book describes algorithmic methods and hardware implementations that aim to help realize the promise of compressed sensing (CS), namely the ability to reconstruct high-dimensional signals from a properly chosen low-dimensional "portrait." The authors describe a design flow and low-resource physical realization of a sensing system, based on CS. They highlight pros and cons of different design choices from a pragmatic point of view, and show how a lightweight and mild, but effective form adaptation to the target signals can be the key to consistent resource saving. The basic principle of the design flow described can be applied to the design of almost any compressed-sensing-based system and has been proven with an extremely diversified set of applications. Many pragmatic aspects of putting a compressed-based system to work are addressed, such as saturation, quantization, leakage, etc.
