Application of Deep Learning in Individual Identification of Radio Frequency Circuits after TID

The individual identification method of the irradiated RF circuits is proposed based on deep learning technology. In order to identify the same RF circuit under different conditions of TID (Total Ionizing Dosem), the deep residual network architecture is built. So that, the recognition results can be used for the research of the TID and damage mechanism of RF circuits, while the actual space exposure times can be estimated. At last, the RF circuits with same type are irradiated with different doses of radiation, and the residual network is used to estimate the actual space exposure years of these circuits, with the accuracy more than 83.92%.