Joint data freshness and throughput optimization in UAV-aided data collection networks under jamming attacks

This paper investigates joint data freshness and throughput optimization in unmanned aerial vehicle (UAV)-aided data collection networks that are exposed to intentional jamming. We first develop a computation-coupled age of information (AoI) model for a UAV–base station (BS) link, where the sensing updates generated by a data collection UAV are transmitted to a remote base station and processed by an edge server. A closed-form expression for the average AoI is derived as a function of the effective throughput and the BS computing rate, thereby explicitly capturing the coupling between information timeliness and transmission rate. To characterize the hierarchical confrontation between a malicious attacker and the UAV, we formulate a Stackelberg game in which the jammer (leader) chooses its jamming power and the UAV (follower) adapts its transmit power to maximize their respective utilities that jointly account for AoI, throughput, and power cost under queue-stability and QoS constraints. The resulting bilevel optimization is solved via a backward-induction framework equipped with a derivative-free golden section search with parabolic interpolation (BI–GSSPI), whose convergence, complexity, and equilibrium properties are analyzed. Simulation results demonstrate that the proposed method significantly improves the utility of the UAV while yielding rational payoffs for the jammer, outperforming average-power and random baselines under various parameter settings.