ED- FreEst: Event - Driven Frequency Estimation

Event-driven or Time-encoded sampling is an alternative to the conventional uniform sampling paradigm that encodes the amplitude information of a continuous-time signal into a sequence of time stamps. The data-driven approach of event-driven acquisition provides significant power efficiency advantages, as it initiates sampling exclusively upon the detection of specific events, such as amplitude changes. In recent years, the focus of event-driven sampling has shifted from bandlimited function classes to time-domain sparse signals. That said, the case of Fourier-domain sparse signals (frequency or spectral estimation problem) remains open. In this paper, we introduce a novel method for off-the-grid, event-driven frequency estimation (ED-FreEst). Empirically, our algorithm results in a lower sampling rate while offering robustness. These aspects seamlessly translate into real-world validation. To demonstrate this, we build an event-driven sampling hardware utilizing asynchronous sigma-delta modulators, showcasing the practical utility and effectiveness of our method in tangible applications.