Radar radiation source identification method and device based on extended residual network

Abstract

The invention belongs to the technical field of radar signal processing, and particularly relates to a radar radiation source identification method and device based on an extended residual network. The radar radiation source identification method comprises the steps: carrying out the time-frequency analysis of a radar signal, and converting the time domain waveform of the radar signal into a two-dimensional time-frequency image; preprocessing the time-frequency image to obtain input data of a deep learning network; and constructing an extended residual deep learning network model, and for input data, self-learning signal time-frequency image features by using the network model and carrying out classified identification. According to the radar radiation source identification method, the problems that a traditional method is sensitive to noise, low in extraction characteristic effectiveness and universality and the like are solved, and the excellent identification effect can still be kept for complex multi-class radar signals in the environment with the low signal-to-noise ratio; the radar radiation source identification method can solve the problems that a simple depth model is weakin learning ability, confusion time-frequency image similar signals and the like, and is good in confusion resistance, accurate in identification result and high in identification accuracy; and the radar radiation source identification method can be applied to radar radiation source identification of more types, and has very high adaptability and popularization.

Description

technical field [0001] The invention belongs to the technical field of radar signal processing, in particular to a radar radiation source identification method and device based on an expanded residual network. Background technique [0002] With the increasingly complex electromagnetic environment and complex and changeable radar signal patterns, the identification of radar radiation sources is facing severe challenges. The traditional measurement method based on the external characteristics of pulse descriptors can no longer meet the needs of quickly and accurately distinguishing radiation sources. Therefore, research The focus turns to extracting the intrapulse features of the signal, such as intrapulse time domain features, ambiguity function features, and time-frequency domain features. A method that can represent the essential characteristics of different radar signals is needed. Deep learning learns features with more classification capabilities by building a neural ne...

AI Technical Summary

Problems solved by technology

[0003] For this reason, the present invention provides a radar radiation source identification method and device based on an expanded residual network, which overcomes the problems of high signal quality requirements, low effectiveness, and strong pertinence of the manual extraction feature method, and realizes the recognition of radar radiation sources under low signal-to-noise ratios. The classification and identification of complex multi-class radar signals has a strong application prospect

Images