A method for classification and recognition of communication radiation source signals

Abstract

Provides a communication radiation source signal classification and recognition method under Gaussian white noise channel conditions, including: transmitting and receiving communication radiation source signals; extracting envelope information from communication radiation source received signals; constructing communication radiation source signal feature data set dictionary ; Classification and identification of signals received by unlabeled communication radiation sources. On the premise that the tagged signal feature dictionary database is artificially supplemented and improved, the present invention can realize good classification and identification of communication radiation source signals under Gaussian white noise channel conditions.

Description

technical field [0001] The invention relates to wireless communication and signal processing technology, in particular to a method for classifying and identifying communication radiation source signals. Background technique [0002] At present, the commonly used classification and identification methods of communication radiation source signals are: K-nearest neighbor clustering method, which is a commonly used classification method because of its simplicity, effectiveness, and easy implementation. However, its classification performance is affected by the neighbor parameter K, the similarity measurement method of neighbor points, the size and distribution of the data set. Support vector base classification method, support vector base is a pattern classification technique proposed by Vapnic in 1995. This method is based on the VC dimension theory of statistical learning theory and structural risk minimization, and seeks the best compromise between the complexity of the mode...

AI Technical Summary

Problems solved by technology

This method is effective in solving small sample classification problems, but the biggest disadvantage of this method is that when the training signal data is too large, the efficiency is low, and the memory usage is too large. At the same time, the selection of different kernel functions will also lead to classification effects. difference

Images