Chen Lingyu, Su Youxing, Luo Wenbin, Hong Xuemin, Shi Jianghong
School of Information Science and Technology, Xiamen University, Xiamen 361005, China.
Key Lab of Underwater Acoustic Communication and Marine Information Technology, Ministry of Education, Xiamen 361005, China.
Sensors (Basel). 2018 Mar 22;18(4):940. doi: 10.3390/s18040940.
The deployment density and computational power of small base stations (BSs) are expected to increase significantly in the next generation mobile communication networks. These BSs form the mobile edge network, which is a pervasive and distributed infrastructure that can empower a variety of edge/fog computing applications. This paper proposes a novel edge-computing application called explicit caching, which stores selective contents at BSs and exposes such contents to local users for interactive browsing and download. We formulate the explicit caching problem as a joint content recommendation, caching, and delivery problem, which aims to maximize the expected user quality-of-experience (QoE) with varying degrees of cross-layer sensing capability. Optimal and effective heuristic algorithms are presented to solve the problem. The theoretical performance bounds of the explicit caching system are derived in simplified scenarios. The impacts of cache storage space, BS backhaul capacity, cross-layer information, and user mobility on the system performance are simulated and discussed in realistic scenarios. Results suggest that, compared with conventional implicit caching schemes, explicit caching can better exploit the mobile edge network infrastructure for personalized content dissemination.
预计在下一代移动通信网络中，小型基站（BS）的部署密度和计算能力将显著提高。这些基站构成了移动边缘网络，这是一种普遍存在的分布式基础设施，能够支持各种边缘/雾计算应用。本文提出了一种名为显式缓存的新型边缘计算应用，它在基站存储选择性内容，并将这些内容提供给本地用户进行交互式浏览和下载。我们将显式缓存问题表述为一个联合内容推荐、缓存和传输问题，旨在在不同程度的跨层感知能力下最大化预期用户体验质量（QoE）。提出了最优且有效的启发式算法来解决该问题。在简化场景中推导了显式缓存系统的理论性能界限。在实际场景中模拟并讨论了缓存存储空间、基站回程容量、跨层信息和用户移动性对系统性能的影响。结果表明，与传统的隐式缓存方案相比，显式缓存能够更好地利用移动边缘网络基础设施进行个性化内容分发。
