Precise: Predictive Content Dissemination Schemes Exploiting Realistic Mobility Patterns

Abstract

Device-to-Device (D2D) communications have expanded the way of managing available network resources to efficiently distribute data between users. D2D exploits communication alternatives, in Opportunistic Networks, based on short range wireless radio technologies such as Bluetooth and WiFi-Direct. Besides, nowadays in most urban areas, realistic human mobility is characterized by often repeated patterns that can be used to accurately predict the next visited regions—we call these regions hotspots (or Replication Zones (RZs)). In this work, we present Predictive Content Dissemination Scheme (Precise), to explore and combine the D2D paradigm along with real mobility and predictions focused on the dissemination of content among hotspots. To analyze the viability of such scheme, we show simulation results and evaluate the average content availability, lifetime and delivery delay, storage usage and network utilization metrics. We compare the performance of Precise with state-of-the-art approaches, such as Epidemic, restricted Epidemic, and Proximity-Interest-Social (PIS) routing protocols. Our results underline the need for smart usage of communication opportunities and storage. We demonstrate that Precise allows for a neat reduction in network activity by decreasing the number of data exchanges by up to 92%, requiring the use of up to 50% less of on-device storage. This comes at negligible costs. In particular, the delivery delay with Precise shows an increase with respect to epidemic dissemination schemes that varies from 0.03 seconds in the most dynamic case to at most 1.91 seconds for the least dynamic case, and which however does not hinder the possibility to use Precise for real-time applications. Regarding how contents are spread, we observe that Precise requires 2% to 20% less mobile users to carry them within a target hotspot, especially under slow dynamics. This however does not impact on the probability that mobile users entering the hotspots obtain contents, and barely shortens the lifetime of contents in our experiments from 100 minutes down to about 95, in the worst case. This demonstrates that the reduction of content availability among mobile users with Precise is either negligible or not impactful, thus guaranteeing the dissemination of contents as with legacy epidemic dissemination protocols.