Blockage-Aware Anti-Jamming Data Harvesting for UAV-Assisted Communications
Abstract
This paper investigates a blockage-aware anti-jamming strategy for UAV-assisted data harvesting in dense urban environments. Ground jammers (GJs) not only emit interference but also attempt to detect UAVs via line-of-sight (LoS) surveillance. To counter this, the UAV leverages urban building layouts to remain hidden from GJs and attenuate their jamming signals, while maintaining LoS connectivity with ground sensors (GSs) for secure and efficient data collection. To maximize the minimum average spectral efficiency among GSs, we jointly optimize the user scheduling and UAV trajectory based on a blockage-aware channel model that captures LoS/NLoS conditions induced by building layouts and relative node positions. This results in a challenging mixed-integer nonconvex problem due to binary scheduling variables, nonconvex trajectory constraints, and the use of sigmoid-approximated channel models. To tackle this, we decompose the original problem into two convex subproblems using successive convex approximation, quadratic transform, and geometry-based convexification techniques. An iterative block coordinate descent-based algorithm is then proposed, enabling efficient convergence with polynomial complexity. Simulation results confirm the superiority of the proposed scheme over baselines, particularly under severe jamming and building blockage conditions, demonstrating improved resilience and throughput in urban UAV networks.