Radiation source signal identification method of improved particle swarm extreme learning machine

Abstract

The invention discloses a radiation source signal identification method of an improved particle swarm extreme learning machine. The method comprises the following steps of S1, preprocessing a radiation source signal; S2, extracting characteristic parameters of the preprocessed signal obtained in the step S1, and obtaining a training sample and a test sample; S3, putting the training sample obtained in the step S2 into an extreme learning machine, initializing parameters of a particle swarm optimization algorithm, and obtaining a learning factor value by an exponential function method; and S4, through learning of the extreme learning machine in the step S3, calculating a mean square error as a moderate value and inertia weight division, continuously updating the speed and the position of the particles, and adjusting the connection weight and a threshold value of the extreme learning machine. The method mainly solves a problem that a traditional optimized extreme learning machine is not high in classification precision and the like, fast optimization is carried out in the environment with the low signal-to-noise ratio, the recognition rate reaches 95% or above, fast optimization can be carried out, learning efficiency is improved, and accuracy of individual recognition is improved.

Description

technical field [0001] The invention relates to a radiation source signal identification method for an improved particle swarm extreme learning machine, which can be used in the technical field of signal processing. Background technique [0002] Radiation source individual identification technology, also known as radiation source "fingerprint" identification or specific radiation source identification (Specific Emitter Identification, SEI), refers to the characteristic measurement of the received electromagnetic signal, and the determination of the radiation that generates the signal based on the existing prior information. source individual. Because this technology has a very broad application prospect but has no ready-made theoretical support, it has gradually become a research hotspot and difficulty in the field of electronic reconnaissance. However, there are still some defects in the recognition classifier, such as the low recognition rate of the classifier, training T...

AI Technical Summary

Problems solved by technology

extreme learning machine [5] (ELM) mainly uses randomly generated connection weights and thresholds for classification and recognition, but cannot find the best network structure, resulting in low accuracy

When there are many training samples, the optimized extreme learning machine will also have problems such as long learning time and general optimization effect, which is difficult to meet the needs of practical applications.

[0004] The classifiers based on radiation source signal target recognition that have been proposed so far are: Xu Yulong, Wang Jinming and others published the paper "Radiation Source Fingerprint Feature Extraction Method Based on Wavelet Entropy" published on "Data Acquisition and Processing" in 2014. The network is used as a classifier, and the recognition rate reaches more than 95% only in the environment of 20dB, but the recognition rate cannot meet the classification requirements in the environment of low signal-to-noise ratio

Images