當基地台無法提供服務的時候，為維持該區域的通訊狀況，需要盡快的恢復通訊服務，得以對外取得連結。而大量的使用者湧入某個基地台的服務範圍內的時候，導致的基地台的負載不均問題，可能造成某些使用者沒有辦法獲得通訊的服務。 而環境中負載不平衡的問題，存在於基地台與流動使用者之間的位置。當使用者不均勻地分散在環境中會使得負載不平衡，在流動的使用者中有可能讓負載不均的問題更加嚴重。 在本研究裡，我們透過派遣可在空中移動的無人機作為通訊基地台，無人機可以在短時間內快速派遣，克服地形崎嶇的限制和避免進入災區可能再度造成的危險。以使用者的位置為考量配置無人機基地台提供通訊服務，並透過機器學習的方式動態調整無人機的位置，希望能改善在環境中使用者的位置不均勻分散造成某些基地台的負載過重，以達成負載平衡改善的目的。 根據模擬結果，本文提出的動態調整無人機基地台的方法能改善負載不平衡的問題。;When the base station is unable to provide services, in order to maintain the communication status of the area, it is necessary to restore the communication service as soon as possible, so that the link can be obtained externally.When a large number of users flood into the service area of a certain base station, the uneven load on the base station may cause some users to have no way to obtain communication services. The problem of load imbalance in the environment exists between the base station and the mobile user.When the user is unevenly dispersed in the environment, the load is unbalanced, and the problem of uneven load may be more serious among the flowing users. In this study, we dispatched drones that can be moved in the air as communication base stations. UAVs can be dispatched quickly in a short period of time, overcoming the rugged terrain and avoiding the dangers that may arise again in the disaster area.The UAV base station is configured to provide communication services based on the user′s location, and the position of the UAV is dynamically adjusted through machine learning. It is hoped that the uneven distribution of users in the environment will cause the base station to be overloaded.To achieve the goal of load balancing improvement. According to the simulation results, the method of dynamically adjusting the UAV base station proposed in this paper can improve the load imbalance.
