Intelligent emitter identification method based on GRU depth convolution network

Abstract

The invention relates to an intelligent emitter identification method based on GRU depth convolution network, which mainly solves the problem that the serialization characteristics of radar emitter signals can not be extracted by the prior art. The scheme comprises the following steps: radar emitter signals are classified; the radar emitter signals are simulated and sliced, the sliced samples areconverted into two-dimensional real samples, and the two-dimensional real samples are normalized and divided into training sample set and test sample set, a depth neural network based on gated circulation unit GRU is constructed; the training sample set is inputted into the depth neural network, and the trained depth neural network model is obtained by optimizing the loss function, the test sampleset is input into the trained depth neural network model, and the recognition result of radar emitter signal is obtained. The invention can extract pre-and post-correlation features of signals, avoids manual feature extraction and prior knowledge, has low complexity and accurate classification result, and can be used for radar emitter identification under complex electromagnetic environment.

Description

technical field [0001] The invention belongs to the technical field of radar, and further relates to a radiation source signal identification method, which can be used for automatic feature extraction and related parameter identification of radar radiation source signals in complex and changeable electromagnetic environments. Background technique [0002] Radar emitter signal classification and identification is an important part of modern electronic intelligence reconnaissance system and electronic support system, and it is of great significance to national defense construction. After years of painstaking research, the identification of radar emitter signals has made great progress. The traditional signal identification method based on five radar parameter characteristics is no longer suitable for modern electronic warfare. Therefore, some scholars extract intrapulse feature information from radar radiation sources for identification, and can obtain a satisfactory accuracy...

AI Technical Summary

Problems solved by technology

This method can achieve a high recognition rate in the identification of radiation source signals, solve the shortcomings of traditional identification methods that are highly dependent on prior knowledge, and complete the extraction of relevant features before and after the signal, so as to identify radiation sources in more complex situations. signal for identification

Images