您当前访问的的浏览器版本过低,为了给您带来更好的体验,建议您升级至Edge浏览器或者推荐使用Google浏览器
取消

一种基于公平性的无人机基站通信智能资源调度方法

作者:吴官翰,赵建伟,高飞飞 阅读量:993

一种基于公平性的无人机基站通信智能资源调度方法

吴官翰1,赵建伟1,高飞飞2
(1. 火箭军工程大学,中国 西安710038;2. 清华大学,中国 北京 100084 )

摘要:空天地一体化网络是未来6G的关键内容。结合高精度波束赋形,无人机(UAV)的视距链路(LoS)可很好地作为空天地一体化网络的补充,但地面用户与基站间的相对运动极易造成信道容量失衡。提出一种噪声深度确定性策略梯度(Noisy-DDPG)方法,该方法以最大化通信公平性和系统容量为目标,利用DDPG优化分配方案,通过调整奖励函数策略参数来实现公平性和信道容量的平衡;通过在策略网络中利用可学习参数噪声进行扰动,得到更合理的分配方案。仿真实验验证了该算法的有效性。  
关键词:无人机基站;资源调度;DDPG;公平通信;参数噪声  


Intelligent Resource Allocation Based on Fairness for UAV Base Station Communications

WU Guanhan1, ZHAO Jianwei1, GAO Feifei2
(1. Rocket Force University of Engineering, Xi’an 710038, China; 2. Tsinghua University, Beijing 100084, China " )

Abstract: The space-air-ground integrated network is an important part of the future 6G, which can be well complemented by the unmanned aerial vehicle’s (UAV) line-of-sight (LoS) link combined with high-precision beamforming. However, the random channel characteristics of mobile users can easily cause channel capacity imbalance. In this paper, the Noisy-Deep Deterministic Policy Gradient (DDPG) is proposed. To maximize communication fairness and system capacity, the deep deterministic policy gradient (DDPG) is used to optimize the allocation strategy. Besides, fairness and channel capacity are differently emphasized by adjusting the reward function policy parameters. Moreover, the learnable parameter noise is used to disturb the policy network to obtain a more reasonable allocation plan. Finally, various simulation results to verify the effectiveness of the algorithm are proposed.  
Keywords: UAV base station; resource allocation; DDPG; fair communication; parameter noise

在线PDF浏览:PDF