Assessing the Probability of Detection in a Frequency-Selective EMD Detector: A Design Perspective

Abstract

Electronic devices exposed to complex electromagnetic environments can fail at specific frequencies. Early, frequency-selective detection of such “weak spots” is therefore desirable, whereas time-domain field probes provide only broadband indications and may over-trigger. This article presents a frequency-selective detector based on software-defined radio with field-programmable gate array acceleration, which transforms time-domain samples into spectra with reduced latency. The study focuses on narrowband burst electromagnetic disturbances and develops closed-form expressions for the probability of detection as a function of front-end bandwidth, frequency span, dwell time, hopping time, frequency resolution, and disturbance duration. The analysis is corroborated by Monte Carlo simulation and conducted measurements across 70 MHz–6 GHz. Results reveal how design choices and hardware latencies bound the attainable detection probability and delineate the hopping-limited regime. This article concludes with practical boundary conditions and design guidelines to maximize detection performance under resource constraints, offering a configurable alternative where high-end real-time analysers are impractical.