ENHANCED APPROACH FOR THE IDENTIFICATION OF HAZARDOUS DIGITAL RADIO EMISSIONS

Abstract

The identification and classification of digital radio emissions constitute a significant technical hurdle, especially within electromagnetically contested sectors. This research investigates low-pass filtering techniques, analyzing architectures where the output exhibits either linear or quadratic dependence on the input amplitude. These filtering mechanisms function by coherently aggregating deterministic signal elements while allowing random noise components to accumulate incoherently. This process leads to constructive amplification of signal energy alongside sublinear noise power growth, thereby markedly improving the signal-to-noise ratio (SNR). A rectangular pulse was utilized as a proxy for contemporary digital communication waveforms across both filter types. A thorough statistical analysis was conducted in both time and frequency domains, evaluating metrics such as variance, mathematical expectation, correlation coefficients, and SNR. To objectively assess filtering efficiency, a new metric termed the "payoff coefficient" was established to measure gains in detection reliability. Simulations further explored envelope voltage responses from an ideal bandpass filter stimulated by rectangular pulses of various durations, mimicking Low-Probability-of-Intercept (LPI) systems. The study confirms that covert signals can be effectively isolated using two-dimensional likelihood density estimation. System-level integration of narrowband low-frequency filtering enhanced the noise immunity of airborne digital signal detection by 23%, bolstering operational resilience in hostile environments relevant to electronic warfare and signals intelligence.