Multiple information cognition of interrupted sampling repeater jamming in complex scenes

Abstract

The dense and varied jamming signals in complex electromagnetic environment pose a great threat to radar detection, especially coherent repeater jamming affects anti-jamming ability seriously. The sufficient cognition of jamming signals is of great significance for effective countermeasures. This paper proposes a multiple information cognition framework and method for interrupted sampling repeater jamming(ISRJ) in complex scenes. Firstly, two one-dimensional convolutional neural networks(1D-CNN) are used to identify jamming types and repeater strategies hierarchically. The various jamming types in complex scenes are identified coarsely, and the specific repeater strategies of ISRJ are identified in detail. Secondly, a one-dimensional parameter estimation deep complex-valued convolutional network (PE-DCCN) is applied to estimate the slice width of ISRJ. Simulation results prove the proposed method can obtain multiple cognition results of ISRJ with high recognition precision and low estimation error.