H2NDN: Supporting Connected Vehicle Applications with Hierarchical Hyperbolic NDN

Abstract

Connected vehicle (CV) has emerged as a new computing and networking paradigm recently that can enable many beneficial CV applications. In this paper, we argue that Named Data Networking (NDN) presents a good synergy with the needs of CV applications. This drives us to propose an NDN based CV application framework with a distributed data service model. However, the fast mobility and vast moving area of vehicles lead to significant challenges on the efficiency and scalability of the underlying NDN backbone in the framework. To handle this issue, we propose a novel hierarchical hyperbolic NDN architecture (H2NDN). By taking advantage of the location dependency of CV applications, we design a hierarchical router topology and the associated namespace. This enables Interest packets from CV applications to be forwarded towards Data (cache) holders efficiently through static FIB configuration. However, the resultant hierarchical routing can easily overload high-level routers. Thus, we further integrate hyperbolic routing into the architecture through carefully designed hyperbolic planes in the hierarchical topology. As a result, the load is better balanced to ensure the overall scalability. Finally, extensive NDNsim based simulation with real traffic data proves the efficiency and scalability of the proposed H2NDN architecture.