Fight Against Smart Communication Rival: An Intelligent Jamming Approach With Trend-Oriented Efficacy Evaluation

• DOI: 10.1109/lwc.2022.3199743
• Publication Date: 2022-08-25T18:37:16Z
• Authors: Zhibin Feng; Yuhua Xu; Yutao Jiao; Guoxin Li; Wen Li; Gui Fang
• Keywords: 0203 mechanical engineering; 02 engineering and technology

This letter investigates a jamming problem in the confrontation scenario when both opposites are intelligent. Most existing studies assumed learning-based jammer could obtain communication feedback as reward, which may be infeasible and limits application. To overcome this limitation, we propose efficacy evaluation method based on change of behaviors, is oriented by idea to reflect trend real effect approximatively. Considering rival intelligent adjust strategy dynamically, multi deep-learning classifiers established predict channel status make selectively, aiming disturb rival's normal communication. Moreover, power optimized multi-armed bandit (MAB) within energy constraint. Simulation results demonstrate effectiveness criteria show that proposed algorithm can decrease throughput well saving cost.