The Research On Identification Technology Of Communication Emitters And Key Nodes

Modulation recognition,communication emitter identification,and key nodes identification are key links in electronic reconnaissance missions for UAV swarms.In an increasingly complex environment,the requirements for feature extraction in communication radiation source modulation recognition and communication emitter identification are getting higher and higher.As the scale of the UAV cluster network becomes larger and the dynamic changes of the communication network increase,the difficulty of identifying key nodes in the dynamic timing network is gradually increasing.This thesis focuses on communication signal modulation recognition,communication emitter identification and key nodes identification based on network topology.The main work of this thesis is as follows:In terms of modulation recognition,in order to solve the problem of low recognition accuracy when the signal modulation types are complex,the simulation data set containing 9 digital modulation signals is used for modulation recognition by using the feature vector combined with the high-order cumulant characteristic parameters in both time domain and frequency domain.The recognition task achieves better performance under the non-Rician multipath channel condition.Aiming at solving the problem that the signal is affected by Rician multipath channel and other undesirable factors,and the traditional method cannot meet the recognition requirements when the number of signal categories is greater,the long short-term memory(LSTM)network is used to carry out feature extraction and recognition towards the data set from real world containing 12 kinds of digital modulation signals.The experimental results show that the LSTM method is more suitable for the modulation recognition task towards data sets from real world with more complex modulation types,and is less susceptible to the influence of Rician multipath channel.In terms of communication emitter identification,in view of the problem that the recognition accuracy is not high when the radiation source signal has bit similarity,a large number of signal categories,and when the length of the collected signal is limited,three feature extraction methods,including two integral bispectrals features and a empiricalwavelet-transform edge spectrum feature,are used when the communication emitter identification task is performed on the bit-similar communication signals from 16 USRPs of the same model at different transceiver distances,and the overlapping sub-sequence division method was proposed as a means of data enhancement.The experimental results show that when the signal sample length is fixed,the training set generated by using the overlapping sub-sequence division method can better train the model than the nonoverlapping sub-sequence division method,which significantly improves the recognition accuracy of the model.In terms of key nodes identification,considering the impact of the dynamic timing of UAV swarm ad hoc network on the static key node identification method,in order to improve the effectiveness of dynamic attacks on UAV key nodes,based on the betweenness centrality index,A key nodes identification method based on an entropyweighted betweenness centrality index is proposed.Experimental results show that this method can improve the robustness of key nodes identification in dynamic networks without increasing the perception time.