NOMA-MEC Based Task Offloading Algorithm in UAV-Assisted IoV Networks

In recent years, the development of B5G/6G technologies has significantly increased the demand for low-latency and low-energy devices in telematics scenarios. However, the limited computing power of mobile devices places them under tremendous pressure when performing computationally intensive tasks. Non-orthogonal multiple access (NOMA) technology, which allows multiple users to share the same resource elements in time, frequency, space, and code domains, effectively addresses the issue of insufficient computing power in mobile devices. In this context, we propose a NOMA-based UAV-assisted mobile edge computing (MEC) network to enhance overall system performance. This network introduces a UAV-assisted IoV task offloading algorithm (TOA) in a real vehicular networking environment, fully considering the computational power of smart vehicles and the flight speed of UAVs. The aim is to minimize the total task processing time of the system under constraints of computational resources and energy consumption. Specifically, the TOA is transformed into binary encoding, integrated with the traditional genetic algorithm (GA), and the task processing time is used as the fitness value to seek the optimal task processing strategy through an iterative process. Simulation results show that, compared to traditional task allocation methods, TOA can effectively complete the allocation and processing of all tasks while satisfying computational resources and energy consumption constraints, reducing the total task processing time by up to 56%. These results provide crucial support for future research and development of NOMA-MEC in IoT technology.