Coded Caching for Reliable Map Dissemination in Symbiotic Communication Aided Emergency UAV Systems

Abstract

Many rescue missions critically rely on effective perception and real-time decision making, which hinge on efficient data collection and dissemination. This study introduces a hierarchical coded caching framework leveraging symbiotic communication, enabling backscatter devices (BDs) to relay information. By strategically caching coded map fragments across unmanned aerial vehicles (UAVs) equipped with BDs, collaborative uploading is fostered to enhance dissemination reliability. We derive the precise successful transmission probabilities for both symbiotic and active communication modes. A coupling relationship between reliable transmission and effective coverage is established, and a joint probability model is deduced to evaluate the effective recovery area of disaster maps. To maximize effective coverage, a deep reinforcement learning (DRL)-based multi-agent dynamic decision algorithm is proposed, optimizing the selection of sensing UAVs, allocation of bandwidth resources, and adjustment of coded caching parameters to accommodate real-time map updates. Simulation results validate the significant improvement in the effective recovery area of disaster maps achieved by the proposed scheme.