A Review of Non-Uniform Sampling Schemes for Power-Efficient Data Acquisition Systems [Feature]

Non-uniform sampling techniques enhance the efficiency of data acquisition systems by operating at a sub-Nyquist sampling rate while maintaining a comparable output quality. These techniques aid in building data acquisition systems that are aware of the signal characteristics so that the limited power budget can be consumed only on certain valuable sampling points at certain signal events, rather than a fixed set of points uniformly sampled at the Nyquist rate. This paper provides a tutorial review of various proposed non-uniform sampling schemes detailing their underlying mechanisms, potential analog circuitry implementations, and the impact of non-idealities on their performance. The paper presents a comprehensive performance comparison between these methods focusing on key metrics such as power consumption, accuracy, and design complexity. A thorough comparison is achieved through analysis of reported performance in the literature and the conducting of simulations. This review aims to guide readers on choosing the appropriate non-uniform sampling scheme that best fits the application requirements, and on their analog implementations and limitations.