Evolutionary Scheduler for Content Pre-Fetching in Mobile Networks

Recently, an increasing number of mobile users are eagerly using the cellular network in data applications. In particular, multimedia downloads generated by Internet-capable smart phones and other portable devices has been widely recognized as the major source for strains in cellular networks, to a degree where service quality for all users is significantly impacted. In this paper we explore the novel concept of proactive content caching using evolutionary algorithms inspired by the inherent predictability of the mobile user behavior. Users can then use the cached version of the content in order to achieve a better user experience and reduce the peak-to-average ratio in mobile networks, especially during peak hours of the day. Finally, we confirm the merits of the proposed scheduler using real data traces of different user's requests and Wi-Fi availability. The results after applying the proposed scheduling algorithm show that up to 70% of the user content requests can be fulfilled i.e. the content were successfully cached before request. We also observe that proposed scheduler outperforms a baseline scheduler based on simulated annealing.