This paper aims at the problems of difficult mobility prediction, rigid topology and resource conflict faced by dynamic path planning in unmanned aerial vehicle-assisted mobile edge computing, and proposes a trinity framework integrating environmental perception, topology optimization and resource collaboration. The accuracy of equipment trajectory prediction is improved through the terrain coupling displacement model, the time-varying weight function is designed to achieve dynamic network reconstruction, and the three-dimensional optimization model of energy consumption-delay-safety is established. Experiments show that this framework reduces the average task delay to 4.2ms (96.5% of tasks <13ms), the delay increase under attack is only 10.5%, and the dispersion of service node distribution is reduced by 42%, providing a theoretically complete solution for the 6G air-space-ground cooperative network.
Zhang et al. (Mon,) studied this question.