Dynamic femtocaching for mobile users

Femtocaching is a caching system to assist the popular content downloading services in heterogenous networks, in which femto base stations (FBSs) utilize their storage capabilities to cache popular files for mobile users (MUs). When the requested files are cached, the content can be downloaded directly from the FBSs through high-rate wireless links, which avoids the backhaul bottleneck to the core network. Previous studies focus on the optimal caching strategy for a given network topology, which is referred to as static femtocaching. However, due to the mobility of MUs, the topology of a practical network rarely stays unchanged and the FBSs need periodically refreshing their caches to adapt to the current network. Limited by the weak backhaul of FBSs, the cache refreshing rate may not catch up with the changing topology, which makes dynamic femtocaching essentially different from the static scenario. In this paper, we first formulate dynamic femtocaching as an optimization problem which is proved to be NP-hard. Then, we propose two dynamic algorithms, centralized and decentralized, to give suboptimal solutions. Simulation results show that the mobility of MUs degrades the performance of dynamic femtocaching for all algorithms, while the proposed algorithms perform 18% ~ 24% better than the traditional algorithm proposed for static scenarios, and 22% ~ 25% better than a simple popular caching system.