This is the authors' abstract. We don't add key points for this paper.
Specific emitter identification (SEI) is crucial for providing operational intelligence in electronic intelligence (ELINT) systems. However, specific operational parameters are required to achieve SEI. With cognitive radar technology advancements, the diversity and uncertainty of signal operational parameters present significant challenges for the representation and identification of radio frequency fingerprints (RFF). Therefore, this study proposes a two-stage metric identification method to better address SEI requirements in ELINT for variable operational parameters (VOP). First, a multidomain-metric loss function is introduced to train a metric model by leveraging deep metric learning theory, which enables the exploration of a metric-feature domain. This method generates metric-RFF (M-RFF) features for each emitter class, thereby overcoming the limitations of traditional classification methods. A lightweight metric identification strategy is employed to facilitate rapid metric decisions using metric score computation based on the M-RFF representation strategy, thereby utilizing the advantages of independent metric concepts. Experimental results demonstrate that the proposed metric identification method outperforms state-of-the-art methods in VOP-SEI tasks. This effectively mitigates the adverse effects of parameter variability on identification performance. This research advances the application of independent metric identification theory to VOP-SEI tasks.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: