Named data networking in vehicular environment: forwarding and mobility
Vehicular networking has become an attractive scenario, where Intelligent Transportation Systems can provide tremendous benefits. Realizing this vision needs a careful design of network architecture due to the dynamic vehicle mobility, unbalanced network density, and specific requirements of its application. The current dominant Internet Protocol (IP) is challenged in this vehicular environment. Named Data Networking (NDN), a proposed future Internet architecture, is more suitable. In this paradigm, interest forwarding is challenging. Further, vehicle mobility plays a significant role in sustaining NDN. Therefore, this dissertation focuses on two aspects of vehicular NDN: interest forwarding and vehicle mobility. The thesis aims to solve the issues of interest forwarding in both dense and sparse networks. Interest forwarding in vehicular NDN suffers packet loss and bandwidth overuse due to broadcast storm, especially in dense networks. Our proposed work, Location-Based Deferred Broadcast (LBDB) scheme, takes advantage of location information to set a rebroadcast deferred timer before rebroadcast. Intermittent connectivity in sparse networks leads to undeliverable packets and long response delay in vehicular NDN. We thus propose a Density-Aware Delay-Tolerant (DADT) interest forwarding strategy that uses directional network density to make retransmission decisions. We have fully implemented LBDB and DADT in simulations. Our evaluation results show that they outperform other protocols for the desired metrics. The sustainability of vehicular NDN is highly related to vehicle mobility, which is influenced by traffic signal operations in urban areas. Our work employs an empirical approach to study the impact from the coordination of traffic signal operations on the capacity and persistence of a vehicle-crowd (v-crowd) that is connected via vehicular NDN. The work delivers practical guidelines for adjusting signals in terms of desired capacity of v-crowd.