Effective gateway association schemes for performance improvement of V2I communications in heterogeneous vehicular networks

Vehicular communication is one of the key enabling technologies for the emerging intelligent transportation systems, which generally consists of direct links between vehicles in proximity (vehicle-to-vehicle, V2V) and roadside infrastructure (vehicle-to-infrastructure, V2I). However, due to the high velocity of vehicles, the throughput of V2I links can be highly dynamic for vehicles passing through different areas. To improve the performance of V2I communications, vehicles with strong V2I links are selected as mobile gateways to relay the data of vehicles in dark areas via V2V communications. In this paper, we study the gateway association problem, for which we formulate a multiple knapsack problem with assignment restrictions to model our optimization task and propose two distributed association schemes to maximize the number of vehicles served by mobile gateways. Simulation results show that our proposed schemes out-perform the distance-based and the random association schemes in various scenarios with different traffic densities.