Dual-Mode Anti-Jamming Strategy for UAV Communications

This paper proposes a novel dual-mode anti-jamming framework that synergistically integrates intelligent frequency-hopping (IFH) with jamming-based backscatter communication (JBC) to address critical vulnerabilities in unmanned aerial vehicle (UAV) communications. Unlike conventional passive anti-jamming approaches, our solution dynamically switches between jamming avoidance and active jamming signal exploitation, repurposing interference as covert carriers when frequency-hopping spread spectrum (FHSS) efficiency degrades. The framework addresses two core challenges: real-time task-based mode selection and dynamic condition-adaptive channel strategy optimization. We formulate this joint optimization problem as a deep reinforcement learning task and introduce a transfer learning-enhanced DQN mechanism that establishes shared latent representations between operational modes, achieving faster convergence than baseline methods while reducing computational overhead for UAV resource constraints. Experimental results demonstrate superior performance in three key dimensions: resilience against high-intensity barrage jamming (steady-state value = 1.0), stability during jamming mode transitions, and adaptability to reactive jammers.