Hierarchical Caching for Statistical QoS Guaranteed Multimedia Transmissions over 5G Edge Computing Mobile Wireless Networks

Abstract

The 5G mobile wireless networks are expected to provision distinct delay-bounded QoS guarantees for a wide range of multimedia services, applications, and users with extremely diverse requirements. On the other hand, how to efficiently support multimedia services over 5G wireless networks has imposed many new challenging issues not encountered before in 4G wireless networks. Bringing data contents closer to mobile users, the caching-based content- centric edge computing network is a promising candidate network architecture and technique to efficiently guarantee QoS for time-sensitive multimedia transmissions over 5G mobile wireless networks. However, how to collaboratively integrate in-network caching and edge computing for multimedia transmissions over 5G wireless networks has been neither well understood nor thoroughly studied. To remedy these difficulties, in this article we propose hierarchical-caching-based content-centric network architectures and their three different implementation/ control mechanisms over 5G edge computing mobile multimedia wireless networks, where popular multimedia data contents can be cached at different network tiers (e.g., routers, base stations, and mobile devices) to guarantee the statistical delay-bounded QoS for multimedia transmissions while minimizing redundant transmissions. We propose three hierarchical edge caching mechanisms: random hierarchical caching, which maximizes the average effective capacity based on the users' request frequency (i.e., data popularity); proactive hierarchical caching, which maximizes the cache hitting rate by predicting the popularity of data contents and caching the data contents with high popularity nearby mobile users; and game-theory-based hierarchical caching, where three network caching tiers are formulated as three game players in a cooperative game to maximize their aggregate effective capacity under the constraints of caching expenses at each tier. We develop three concrete algorithms to implement and control the three hierarchical caching mechanisms, respectively, for statistical delay-bounded QoS guaranteed multimedia transmissions. Using simulations and numerical analyses, we show that our proposed three hierarchical caching mechanisms significantly outperform other existing schemes in supporting the statistical delay-bounded QoS provisioning over 5G edge computing mobile wireless networks.