Trajectory Optimization of Multiple Urban Air Mobility for Reliable Communications with Integrated Space-Air-Ground Network

In smart cities where the supply of personal vehicles and buildings is intensively constructed, overload and traffic jams sometimes occur. At this time, Urban Air Mobility (UAM) can provide unmanned services such as delivery or taxis regardless of ground infrastructure. However, if there is a problem with the remote control or communication of UAM, there is a serious risk that can cause a lot of casualties immediately. Therefore, optimizing the trajectory of UAMs to avoid collisions between UAMs and ensure reliable communication is necessary. On the other hand, communication through satellites such as Starlink has the advantage of having wide coverage, so satellites can support communication services in areas where communication infrastructure is insufficient. Thus, in this paper, reliable communication of UAMs is funded through the satellite-air-terrestrial integrated network. At this time, each UAM considers the location of the communication infrastructure and other UAMs, avoids collisions, and derives a trajectory for reliable communication. We proposed an algorithm based on proximal policy optimization (PPO) to derive the optimal trajectory of UAMs in a short time. As a result of the experiment, UAMs can avoid collisions with other UAMs and simultaneously receive a reliable communication service.