A Signal Clustering Method

Abstract

The invention relates to a signal clustering method in the field of electromagnetic environment signal monitoring. First, the receiver reads the pulse description word PDW i , and then according to the pulse description word PDW i Doa i , carrier frequency Rf i , calculate the PDW i The index address of the PDW; i Judgment of the content stored in the index address, if the orientation Doa i If the next neighbor index address does not have a channel number mark, it will detect Doa in the same direction i The carrier frequency Rf i Upper neighbor index address; judge the content stored in the index address, judge whether there is a channel number sign, if the carrier frequency Rf i If the upper neighbor index address has no channel number mark, it will detect Doa in the same direction i The carrier frequency Rf i Next-neighbor index address; the present invention provides the method of index address hash table, which can cluster the detection and reception signals in real time and quickly to ensure 100% acquisition of signals; also applies the direction Doa i and carrier frequency Rf i The upper and lower neighbor associations make the signal carrier not split due to movement or the frequency agility of the radar signal, which provides a guarantee for the subsequent signal processing to correctly sort the signal.

Description

technical field [0001] The invention relates to a signal clustering method in the field of electromagnetic environment signal monitoring. Background technique [0002] Due to the dense, complex and changeable modern electromagnetic environment, the received signals are very dense. If these signals are not classified and diluted, it is difficult to quickly separate the signals, especially high repetition frequency signals and low repetition frequency signals. When the signals are mixed together, the low-repetition-frequency signals will be submerged and cannot be sorted out; in order to adapt to the external pulse density of 1 million per second, the processing time for each pulse should be less than 1us, so that the pulse will not be lost, ensuring 100% signal acquisition; there are many existing clustering methods, such as: neural network, vector machine, Kalman filter, etc. These methods take a lot of time and are difficult to be widely used in engineering. Contents of t...

AI Technical Summary

Problems solved by technology

[0002] Due to the dense, complex and changeable modern electromagnetic environment, the received signals are very dense. If these signals are not classified and diluted, it is difficult to quickly separate the signals, especially high repetition frequency signals and low repetition frequency signals. When the signals are mixed together, the low-repetition-frequency signals will be submerged and cannot be sorted out; in order to adapt to the external pulse density of 1 million per second, the processing time for each pulse should be less than 1us, so that the pulse will not be lost, ensuring 100% signal acquisition; there are many existing clustering methods, such as: neural network, vector machine, Kalman filter, etc. These methods take a lot of time and are difficult to be widely used in engineering

Images