Underwater acoustic communication signal identification method, system and apparatus based on sequence convolutional network

Abstract

The invention discloses an underwater acoustic communication signal identification method, system and apparatus based on a sequence convolutional network. The method comprises: constructing a sequence convolutional network model by adopting a mode of stacking a plurality of residual modules, training the sequence convolutional network model, and inputting the underwater acoustic signal into the trained sequence convolutional network model for signal identification. The network structure adopting the design form of stacking a plurality of residual modules can effectively overcome the problem of network model degradation caused by multi-layer network stacking. After the network is trained, the signal type identification capability of the network is improved, and finally, after the underwater acoustic signal is input into the trained network, the signal type identification capability can be improved, so that the identification of the communication signal is realized.

Description

technical field [0001] The present application relates to the technical field of signal recognition, in particular to a method, system and equipment for underwater acoustic communication signal recognition based on sequential convolutional networks. Background technique [0002] Signal identification is usually viewed as communication signals. The task of automatic modulation identification mainly involves classifying the individual signal variables of modulation schemes to identify the communication mode applied between transceivers in an application scenario. Signal recognition has many applicable scenarios in both military and civilian environments. According to different application requirements, signal recognition can be widely used in non-cooperative fields and network security fields. Military environments include the detection, analysis, and identification of unknown signal modulations as potential sources of enemy communications. [0003] Water has a low absorption...

AI Technical Summary

Problems solved by technology

Although lower frequencies are better for the acoustic wave transmission process, communication resources are greatly limited by the available bandwidth, which may typically range from tens of Hertz to several thousand Hertz, these states make the work of underwater signal modulation identification Extremely challenging, unable to achieve fast identification of underwater acoustic signals

Images