Prediction of the content popularity in the 5G network: Auto-regressive, moving-average and exponential smoothing approaches

Today's mobile users want faster data and more reliable services. The next generation of wireless networks 5G promises to deal with this, and more. In this context, to enable ultra-short response times, fast relocation of service instances between edge nodes and reduce migration time its required to cope with user mobility to guarantee the (QoE). In this new paradigm called 5G cellular systems, the technique of content caching in small base stations (SBS) is considered to be a suitable approach to improve the efficiency and to alleviate the backhaul burden and reduce user perceived latency in wireless content delivery. Proactively serving predictable user demands, via caching at base stations (BS) and users' devices is crucial due to storage limitations, but it requires knowledge about the content popularity distribution, which is often not available in advance. Moreover, local content popularity is subject to fluctuations since mobile users with different interests connect to the caching entity over time. In this paper we focus on the prediction of popularity evolution of video contents. Based on the observation of past solicitations of individual video contents. The popularity prediction in this proactive approach relies on AR (Auto-Regressive), MA (Moving-Average) and Exponential smoothing techniques to complete a proposal caching Algorithm to manage cache decisions.