Research Article
Dynamic Network Access for Multi-UAV Networks: A Cloud-Assisted Learning Algorithm
@INPROCEEDINGS{10.1007/978-3-030-00557-3_36, author={Xiaodu Liu and Yitao Xu and Yanlei Duan and Dianxiong Liu and Zhiyong Du}, title={Dynamic Network Access for Multi-UAV Networks: A Cloud-Assisted Learning Algorithm}, proceedings={Machine Learning and Intelligent Communications. Third International Conference, MLICOM 2018, Hangzhou, China, July 6-8, 2018, Proceedings}, proceedings_a={MLICOM}, year={2018}, month={10}, keywords={Dynamic network access Multi-UAV communication Cloud-assisted SLA}, doi={10.1007/978-3-030-00557-3_36} }- Xiaodu Liu
Yitao Xu
Yanlei Duan
Dianxiong Liu
Zhiyong Du
Year: 2018
Dynamic Network Access for Multi-UAV Networks: A Cloud-Assisted Learning Algorithm
MLICOM
Springer
DOI: 10.1007/978-3-030-00557-3_36
Abstract
In this paper, we study the strategy of UAV dynamic network access in the large-scale UAVs swam. We model the master UAV providing communication coverage for the small UAVs which transformed the large-scale UAVs communication problem into the optimization problem. Compared to the traditional ground user network access, the characteristic of UAV’s mobility have been considered and each UAV have chance to move to any master UAV for better service. We propose a joint optimization for the throughput and flight loss. Due to the limitation of flight loss, the UAVs can not fly to different networks many times for learning. We set up a load aggregator cloud to help the UAVs simulate the results of each decision. We propose a dynamic network access algorithm based on SLA which is proved to achieve stable solutions with dynamic and incomplete information constraint. The simulation results show that this algorithm can converge to the optimal solution. Also, it is shown that the algorithm has strong robustness and can get good utility than other algorithms regardless of how the environment changing.