Impact of Scheduling and Dropping Policies on the Performance of Vehicular Delay-Tolerant Networks

Vehicular Delay-Tolerant Networks (VDTNs) are a disruptive network architecture based on delay-tolerant network paradigm, gathering contributions from opportunistic and cooperative networks, and optical burst switching paradigm. VDTNs assume out-of-band signaling and handle non-real time applications with a low cost network infrastructure. In VDTNs, vehicles are opportunistically exploited to carry data between terminal nodes, enabling network connectivity under unreliable conditions with unstable links and where a contemporaneous end-to-end path may not exist. To address this problem VDTN combines routing schemes that replicate bundles at contact opportunities, with long-term bundle storage. However, this combination increases the resources consumption (e.g., bandwidth, storage) and may affect the performance of the entire network. To improve network performance different scheduling and dropping policies can be used. This paper studies the impact of different scheduling and dropping policies on the performance of a VDTN laboratory testbed using Epidemic and Spray and Wait (binary variant) routing schemes. It was shown that network performance increases, in terms of delivery ratio and delivery delay, when these scheduling and dropping policies are based on the bundle lifetime criteria.