Ince, Ahmet MelihCanbilen, Ayse ElifYanikomeroglu, Halim2026-02-102026-02-1020262644-125Xhttps://doi.org/10.1109/OJCOMS.2026.3651027https://hdl.handle.net/20.500.13091/12972In the dynamic landscape of wireless communication systems, high-altitude platform stations (HAPS) technology heralds a new era of connectivity solutions. The HAPS ensures uninterrupted operation even under challenging conditions where connectivity via terrestrial networks is unavailable. This approach effectively supports real-time applications by dynamically optimizing resource allocation and communication modes. Considering that, this research addresses the strategic integration of HAPS into vehicle-to-everything (V2X) networks. Specifically, multiple autonomous platoons using V2X technology distribute cooperative awareness messages (CAMs) to their followers, attempting to ensure the timely delivery of safety-critical messages not only to the roadside unit (RSU) but also to the HAPS, introducing link-level redundancy to the wireless network. We formulate a multi-objective optimization problem to minimize the age of information (AoI) and power consumption while maximizing the probability of CAM delivery rate. We utilize a multi-agent deep reinforcement learning (MADRL) based resource allocation framework, where each platoon leader (PL) acts as an agent and interacts with the environment to learn its optimal policy. In this framework, based on a deep deterministic policy gradient (DDPG) algorithm, in addition to a local critic trained to predict the individual reward of each PL, a global critic is also trained to predict the global expected reward and motivate PLs to cooperative behavior. The presented simulation results demonstrate the effectiveness of HAPS integration in the considered V2X scenario and the superiority of the proposed algorithm over benchmark algorithms in terms of AoI and power consumption performance.eninfo:eu-repo/semantics/openAccessResource ManagementVehicle-to-EverythingVehicle DynamicsCamsOptimizationPrediction AlgorithmsEnergy ConsumptionDynamic Scheduling6G Mobile CommunicationThroughputAoIHapsResource ManagementReinforcement LearningV2XArticle10.1109/OJCOMS.2026.36510272-s2.0-105028275747