2020/10/29 by Charles E. Thornton, Thornton, Charles E., R. Michael Buehrer +3
Computer Science · Engineering · #FOS: Computer and information sciences #Guidance and Control Systems #Information Theory (cs.IT) #Radar Systems and Signal Processing #Wireless Signal Modulation Classification
paper · pdf · doi:10.48550/arxiv.2010.15698
openalex publication_date 2020/10/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Pulse-agile radar systems have demonstrated favorable performance in dynamic\nelectromagnetic scenarios. However, the use of non-identical waveforms within a\nradar's coherent processing interval may lead to harmful distortion effects\nwhen pulse-Doppler processing is used. This paper presents an online learning\nframework to optimize detection performance while mitigating harmful sidelobe\nlevels. The radar waveform selection process is formulated as a linear\ncontextual bandit problem, within which waveform adaptations which exceed a\ntolerable level of expected distortion are eliminated. The constrained online\nlearning approach is effective and computationally feasible, evidenced by\nsimulations in a radar-communication coexistence scenario and in the presence\nof intentional adaptive jamming. This approach is applied to both stochastic\nand adversarial contextual bandit learning models and the detection performance\nin dynamic scenarios is evaluated.\n