Convolutional network recognition methods for multiple UAVs
Traditional UAV signal recognition methods have some problems,such as low recogni-tion accuracy,poor adaptability to signals in complex environments,and slow speed.The auto-matic recognition method of UAV signal based on convolutional neural network(CNN)is a-dopted to pre-process the real data of UAV signal acquisition,and then a multi-layer model of convolutional neural network is established.The experimental results show that the accuracy rate of all signals is above 90％at 7dB except for royal 2,and the recognition rate of royal 2 reaches 90％at 10dB.Therefore,the use of CNN showed high accuracy and robustness in identifying six types of drone signals.Compared with the previous recognition methods,this method im-proves the recognition accuracy,ADAPTS the signal to the complex environment,and speeds up the recognition speed.
NETL
NSTL
万方数据
