High-frequency pattern mining method of random aggregate radar emitter signal parameters based on weighted clustering

Abstract

The invention discloses a weighted clustering-based method for digging high-frequency patterns of random aggregate radar radiation source signal parameters. Aiming at the signal parameter data of random aggregate airborne radar radiation source targets, a similarity distance matrix is ​​constructed to compare random The similarity between the set measurement values ​​is based on the similarity distance matrix, and iteratively fuses the random set clusters based on the hierarchical clustering method until no cluster that meets the random set distance threshold is found. At the same time, iteratively Calculate the support degree and random set value of each new cluster, and finally dig out the random set value corresponding to the cluster meeting the support threshold as the high-frequency mode output. Advantages of the present invention: (1) Propose a similarity comparison method for random set radar emitter signal parameters; (2) Based on the similarity distance matrix, high-frequency signal parameter patterns can be further excavated; (3) The calculation cost is low , to achieve method engineering.

Description

Technical field [0001] The invention relates to the field of analysis and processing of parameter data of a random set radar radiation source, in particular to a method for mining a high frequency mode of random set radar radiation source signal parameters based on weighted clustering. Background technique [0002] As we all know, there are more and more signal characteristics of today's radar radiators. In addition to traditional types, airborne radar radiators also have a random set of signal parameters composed of several random measurement values. Most of the traditional signal feature types are continuous, and some are discrete. Nowadays, random set signal features have also become an important type of radar radiation source data. This is because, as the complexity of radar radiation sources increases, the types and forms of signal characteristics become more and more diversified. In addition, due to the progress of signal characteristic measurement technology and the infl...

AI Technical Summary

Problems solved by technology

The existing radar emitter signal feature analysis methods and radar emitter type identification methods are almost all for fixed-value signal features, and they cannot describe and analyze random set signal features specifically.

Images