Feature Based on Detection of Aviation Decoy Targets Using Dynamic Radar Cross Section Signatures

Abstract

The proliferation of advanced airborne countermeasure systems has significantly increased the complexity of radar target discrimination in modern aviation environments. Contemporary aircraft deploy towed decoys, chaff clouds, and active radio-frequency repeaters to emulate authentic radar echoes and degrade tracking performance. Conventional amplitude-based radar cross-section (RCS) detection techniques are often insufficient to reliably distinguish genuine aircraft from decoy targets under dynamic flight conditions and low signal-to-noise ratio (SNR) scenarios. This study proposes a robust feature-fusion framework for aviation decoy detection based on dynamic RCS signature analysis across temporal, statistical, and time–frequency domains. Unlike traditional thresholding approaches, the proposed method exploits intrinsic differences in RCS stability, microDoppler structure, spectral dispersion, and entropy characteristics between real aircraft and deployed decoys. In particular, oscillatory scattering behavior associated with towed decoys, stochastic fluctuations characteristic of chaff clouds, and spectral inconsistencies induced by active repeater systems are systematically quantified through multi-domain feature extraction. These features are integrated within a supervised classification architecture optimized for real-time radar processing constraints. A high-fidelity simulation environment representing engagement scenarios involving multirole fighter platforms was developed to evaluate detection robustness under varying SNR levels, aspect angles, and target–decoy separations. Performance assessment using probability of detection, false alarm rate, and receiver operating characteristic (ROC) metrics demonstrates a substantial improvement over conventional amplitude-only RCS discrimination methods. The results confirm that dynamic RCS behavior contains exploitable discriminative signatures that enable reliable identification of aviation decoys without reliance on computationally intensive electromagnetic modeling. The proposed framework provides a scalable and computationally efficient solution suitable for next-generation airborne and ground-based radar systems operating in contested electromagnetic environments.