Content replication and placement in mobile networks

Abstract: Performance and reliability of content access in mobile networks is conditioned by the number and location of content replicas deployed at the network nodes. Location theory has been the traditional, centralized approach to study content replication: computing the number and placement of replicas in a static network can be cast as a facility location problem. The endeavor of this work is to design a practical solution to the above joint optimization problem that is suitable for mobile wireless environments. We thus seek a replication algorithm that is lightweight, distributed, and reactive to network dynamics. We devise a solution that lets nodes (i) share the burden of storing and providing content, so as to achieve load balancing, and (ii) autonomously decide whether to replicate or drop the information, so as to adapt the content availability to dynamic demands and time-varying network topologies. We evaluate our mechanism through simulation, by exploring a wide range of settings, including different node mobility models, content characteristics and system scales. Furthermore, we compare our mechanism to state-of-the-art approaches to content delivery in static and mobile networks. Results show that our mechanism, which uses local measurements only, is: (i) extremely precise in approximating an optimal solution to content placement and replication; (ii) robust against network mobility; (iii) flexible in accommodating various content access patterns. Moreover, our scheme outperforms alternative approaches to content dissemination both in terms of content access delay and access congestion.