Stackelberg-Game-Based Computation Offloading in Urban IoT Systems With AAV-Assisted Multiaccess Edge Computing

Autonomous aerial vehicles (AAV) are regarded as a promising technology to provide additional computing capabilities and wide coverage for Internet of Things (IoT) devices, particularly in cases where these devices are situated beyond the reach of traditional communication infrastructure. This study investigates an AAV-assisted multiaccess edge computing (MEC) network comprising multiple AAVs with edge servers and several IoT Devices (IoTDs). IoTDs with a substantial number of computation tasks can select to offload their tasks to AAV-assisted edge servers to alleviate pressure and costs, while the AAV-assisted edge servers can profit from selling computing resources. The interaction between AAV-assisted edge servers and IoTDs is modeled as a Stackelberg game, where both entities aim to maximize their utility. Employing backward induction, the existence of a unique Nash equilibrium is proved. Subsequently, a Stackelberg game-based distributed computation offloading (SDCO) algorithm is designed to approximate the optimal solution. Finally, extensive simulations validate the effectiveness of the SDCO algorithm, demonstrating superior performance compared to other benchmark methods across diverse scenarios.