A Prototype-Level Study of Jamming-Aware Route-to-Edge Tradeoffs in FANETs

Abstract

Flying ad hoc networks (FANETs) used in counter-UAV missions must make routing decisions under jamming, rapid topology change, and uneven edge workload. Existing studies often treat routing and offloading separately, leaving the route-to-edge tradeoff under adversarial conditions insufficiently examined. This paper presents JCAR-EC, a prototype-level analytical model that jointly scores path quality and edge-processing state, and Robust-JCAR, a revised variant that adds hidden-risk sampling and success-tolerance filtering at the decision stage. The evaluation combines stronger non-learning baselines, paired analytical stress tests, hidden-risk holdouts, a generalized four-candidate edge campaign, and a packet-level ns-3 companion with burst and deadline sweeps. Two findings are central. First, the original JCAR-EC rule is not a uniformly superior heuristic: in the default sensor case it is consistently outperformed by simpler baselines. Second, Robust-JCAR improves the analytical outcomes in the two main positive campaigns. It attains a statistically supported average margin in the hidden-risk holdout and the best average performance in the generalized four-candidate campaign, although the latter advantage over the strongest baselines remains descriptive rather than statistically separated. These findings should be interpreted within the stated scope. The strongest claims are analytical, while the ns-3 layer provides directional packet-level support through a simplified multi-hop companion rather than a full native realization with complete MAC contention and spatial mobility.