摘要
随着无人机(UAV)技术的发展,凭借其高灵活性和快速部署能力,已成为灾后救援场景中不可或缺的工具。针对灾区复杂环境下的通信和能耗优化问题,该研究提出了一种结合动态自适应流媒体编码(DASH)技术和Soft Actor-Critic(SAC)算法的解决方案。通过联合优化视频编码参数、飞行策略及带宽资源分配,以实现在保证用户视频视频体验质量(QoE)的同时,尽可能降低UAV能耗,从而提高系统效益。仿真实验表明,相较于传统深度强化学习(DRL)算法(如DDPG、PPO),该文提出的SAC算法在收敛速度、奖励值及稳定性等方面均具有显著优势。
With the advancement of UAV technology,UAVs have become indispensable tools in post-disaster rescue scenarios due to their high flexibility and rapid deployment capabilities.To address the challenges of communication and energy consumption optimization in complex disaster environments,the study proposes a solution that combines Dynamic Adaptive Streaming over HTTP(DASH)technology and the Soft Actor-Critic(SAC)algorithm.By jointly optimizing video encoding parameters,flight strategies,and bandwidth resource allocation,the proposed method aims to ensure user Quality of Experience(QoE)for video streaming while minimizing UAV energy consumption,thereby enhancing overall system efficiency.Simulation experiments demonstrate that,compared to traditional deep reinforcement learning(DRL)algorithms such as DDPG and PPO,the SAC algorithm proposed in the paper exhibits significant advantages in terms of convergence speed,reward value,and stability.
作者
刘利民
李晋峰
康云鹏
LIU Limin;LI Jinfeng;KANG Yunpeng(China Mobile Communications Group Shanxi Co.,Ltd.Jinzhong Branch,Jinzhong,Shanxi 030600,China)
出处
《长江信息通信》
2025年第9期198-201,210,共5页
Changjiang Information & Communications
关键词
无人机
深度强化学习
灾后救援
SAC算法
路径规划
UAV
Deep Reinforcement Learning
Post-Disaster Rescue
SAC Algorithm
Path Planning
