Design of Cognitive Radar Orthogonal Subpulse Width against Multiple Jamming

This paper develops a Cognitive Radar Orthogonal Subpulse Width Design Framework (CROSWDF) to counter multiple mainlobe Interrupted Sampling Repeater Jamming (ISRJ). Specifically, this framework sequentially emits three subpulses, labeled 1, 2, and 3, in the time domain, where subpulses 1 and 3 are shielding signals designed to deceive the jammers, and subpulse 2 is the probing signal for target detection. The framework employs a greedy search algorithm with a Convolutional Neural Network (CNN) to acquire the Subpulse Widths (SW) of subpulses 1 and 3 online, pulse by pulse, to prevent subpulse 2 from being sampled. The CNN is utilized to check if subpulse 2 is disturbed and to measure the current SW value. Numerical simulations demonstrate that the proposed framework can effectively optimize the SW of subpulses 1 and 3 in the presence of multiple targets and multiple ISRJ with different jamming types and modulated parameters, achieving reliable detection performance compared to some existing methods.