Toward Cooperatively Caching in Multi-UAV-Assisted Network: A Queue-Aware CDS-Based Reinforcement Learning Mechanism With Energy-Efficiency Maximization
Xiaohan Qi, Jingzheng Chong, Qinyu Zhang, Zhihua Yang · IEEE Internet of Things Journal · 2024
With its attractive controllable mobility and flexible deployment advantages, the Unmanned Aerial Vehicle (UAV) has emerged as a promising solution to support temporary caching services by pre-fetching popular content. However, it still exists an obvious challenge due to the limited storage and energy of UAVs with stochastic arrival requests, which is not well addressed by present works resulting in low Quality-of-Service (QoS) for users. In this paper, therefore, we propose a queue-aware cooperatively caching mechanism in the multi-UAV assisted system by considering the random user requests, in which a well-designed Connected Dominating Set (CDS) is developed to make collaborative caching schedule. In particular, we formulate the issue as a long-term queue stability constrained energy efficiency maximization problem by a well-tailored Lyapunov optimization framework. As a non-linear mixed-integer optimization with a nonconvex objective function and coupled variables, we solve it by designing a decentralized Cooperative Reinforcement Learning (CRL) algorithm with the developed CDS. The numerical results demonstrate that our proposed joint algorithm outperforms other benchmark algorithms in terms of caching latency, cache hit ratio, and energy efficiency.