An ultra-low-power analog memory system with an adaptive sampling rate

Sleep states are used in resource-constrained systems to conserve power, but they necessitate a wake-up circuit for detecting unpredictable events. In such systems, all information preceding a wake-up event will be forfeited. In this paper, we present an analog memory system that adaptively samples and records an input signal while the rest of the system sleeps, thereby preserving the information that would otherwise be lost. This system does so while consuming less than 3.52 μW. We also show how the adaptive sampling rate can be utilized to approximate the original signal using a minimal number of samples.