Cognitive Radar Subpulses Waveform Design via Online Greedy Search
Yi Wang, Xianxiang Yu, Jing Yang, Guolong Cui
Abstract
This paper proposes a cognitive radar subpulse waveform design (CRSWD) approach including a smart acquirement process of some a prior information and multi-sequences optimization algorithm against multiple mainlobe interrupted sampling repeater jamming (ISRJ). Specifically, the radar first interacts with the environment on a pulse-by-pulse basis to quest for radar survival window (RSW) dynamically. Hence, a novel RSW searching method incorporating the greedy strategy is devised, where a reasonable value function is defined for measuring the anti-jamming and detection capabilities. Based on the estimated RSW information, orthogonal shielding and probing subpulses are strategically positioned for confusing the jammer and then detecting targets, respectively. To this respect, a non-convex optimization problem based on the peak to-sidelobe level (PSL) criterion and peak-to-average ratio (PAR) constraints is formulated for designing orthogonal probing and shielding waveforms with optimized RSW knowledge. A fast iterative methodology based on block coordinate descent (BCD) and majorize-minimization (MM) framework is proposed with the convergence performance ensured. Numerical simulations demonstrate that the proposed framework can effectively acquire jamming-resistant RSW in the presence of multiple targets and ISRJ with different parameters and achieve reliable detection outperforming some counterparts. Experiments are conducted to further verify its engineering feasibility.