Radar radiation source signal sorting method based on meta learning

Radar radiation source sorting and identification methods based on in-depth learning have received considerable attention in recent years, but there are time and resource constraints in practice. To solve this problem, this paper presents a method of radar radiation source signal sorting based on meta-learning. Data samples for signal sorting are generated by acquiring the three-dimensional entropy, image coefficient, high-order cumulants, bandwidth and carrier frequency characteristics of the signal. A signal classification model based on adaptive channel attention is constructed. Model-Agnostic Meta-Learning (MAML) meta-learning mechanism is used to enhance the generalization performance of the model to meet the task requirements of radar signal sorting under small sample conditions. The experimental results show that this method can effectively improve the low recognition rate of small sample radar radiation source signals, and achieve a good classification accuracy under the condition of a small sample of training data.