Hybrid satellite communication modulation identification method and system based on constellation diagram and deep learning

Abstract

The invention provides a hybrid satellite communication modulation identification method and system based on a constellation diagram and deep learning, and the method comprises the following steps: 1,a receiver obtaining a preset signal through spectral analysis, and carrying out the carrier frequency estimation, bandwidth estimation, band-pass filtering and down-conversion, thereby obtaining a baseband signal; 2, estimating a timing error, and carrying out interpolation filtering on a non-optimal sampling data point to approach an optimal sampling point; 3, performing phase difference on theoptimal sampling point to eliminate the influence of frequency offset, and recovering a constellation diagram after phase difference; and 4, clustering the constellation diagram through a machine learning method, and performing modulation mode classification through features of clustering points. On the basis of a traditional modulation recognition method, fine classification and recognition arecarried out through a mean shift clustering method, the recognition accuracy can be improved through combination of the two methods, meanwhile, the calculation complexity is simplified, and good balance is achieved between the accuracy and the complexity.

Description

technical field [0001] The present invention relates to a hybrid satellite communication modulation identification method and system based on constellation diagram and deep learning, and relates to the field of H04W: wireless communication network. Background technique [0002] Satellite communication is in an open environment and is always facing various intentional or unintentional interferences. Effective means must be used to monitor and counteract these interferences in order to achieve reliable communication. Therefore, it is necessary to further study the modulation recognition technology for satellite communication signals, and gradually establish a modulation recognition system suitable for my country's satellite communication system. [0003] Satellite communication signal recognition technology mainly includes the estimation and identification of parameters such as signal multiple access mode, signal-to-noise ratio, carrier frequency, symbol rate, and modulation m...

AI Technical Summary

Problems solved by technology

[0005] In the traditional method of satellite communication signal modulation identification method, SNR and synchronization have a great influence on the accuracy rate, and the requirements for data length (number of sampling points/number of symbols) are high, and the accuracy rate is low, and the approximate modulation identification method is difficult to separate. The set should not be too large; while the satelli

Images