Efficient task execution is critical in mission-driven Edge-Computing (EC)-based IoT applications such as UAV-assisted weather nowcasting. Although UAV fleets improve the resolution and responsiveness of short-term forecasts, their limited on-board computation and battery life demand a cooperative task offloading strategy. Volunteer vehicular edge networks, leveraging idle onboard units of nearby vehicles, offer scalable, low-cost resources. To incentivize task offloading from UAVs to self-interested vehicles, in this work, we introduce a distributed Dutch auction with second-price payments. The proposed scheme is suited for efficient and time-sensitive allocations. Furthermore, to overcome the challenge of non-trivial task valuation, we embed lightweight Reinforcement Learning (RL) agents that dynamically learn valuation policies. Conducted experiments demonstrate that our approach achieves up to $-93 \%$ in information revelation compared to sealed-bid auctions, up to $-50 \%$ the communication overhead of the best baseline, and linear computation complexity, while preserving high social welfare, which is further improved by the integration of RL.

ABSTRACT Secure and cost‐efficient deployment of Internet of Things (IoT)‐driven business processes (BPs) in multicloud environments is a complex task, particularly under dynamic workloads and strict risk or budgetary constraints. We address this challenge by leveraging deep reinforcement learning (DRL) to determine optimal task‐to‐cloud allocations that comply with user‐defined cost or risk thresholds. Our approach introduces a Q ‐learning model that integrates a confidentiality risk metric and a cost evaluation function into its learning process. Unlike traditional heuristics, the DRL agent adapts to evolving constraints through trial‐and‐error interaction with its environment. Experimental results on a diverse set of deployment configurations show that our approach achieves 25%–30% reductions in both risk and cost compared to heuristic baselines, while satisfying thresholds in 75% of cases. It also demonstrates strong adaptability to dynamic changes, including task additions and resource fluctuations. These findings highlight the potential of reinforcement learning for reliable, constraint‐aware deployment in cloud‐based IoT systems.

International audience