Statistical QoS-Driven Power Allocation for Cooperative Caching over 5G Big Data Mobile Wireless Networks

Abstract

In order to effectively guarantee the statistically delay-bounded multimedia services over time-varying wireless channels, the statistical quality-of-service (QoS) technique has been developed over 5G big data mobile wireless networks. On the other hand, as one of the 5G-promising techniques, the wireless caching enabled WiFi offloading technique is shown to be powerful in addressing the data explosion problem and alleviating the network congestion problem in macrocells. Consequently, challenges have been imposed in applying the cooperative caching schemes for maximizing the successful playback probability under statistical delay-bounded QoS constraints. To effectively overcome the above-mentioned problems, we propose the statistical QoS-driven power allocation scheme through applying the cooperative caching enabled WiFi offloading system over 5G big data mobile wireless networks. In particular, under the Nakagami- \textit{m} fading model, we establish the system models for cooperative caching and wireless transmissions. Given the statistical QoS constraints, we derive and analyze the aggregate effective capacity and the successful playback probability under our developed optimal power allocation policies for the QoS-driven cooperative caching enabled WiFi offloading. Also conducted is a set of simulations which analyze and show the priority of our proposed cooperative caching enabled WiFi offloading scheme, compared with the schemes without cooperative caching in terms of effective capacity under statistical QoS constraints over 5G big data mobile wireless networks.