Geo-localized Virtual Infrastructure for VANETs: Design and Analysis

Supporting future large-scale vehicular networks is expected to require a combination of fixed roadside infrastructure and mobile in-vehicle technologies. The need for an infrastructure, however, considerably decreases the deployment area of VANET applications. In this paper, we propose a self-organizing mechanism to emulate a geo-localized virtual infrastructure (GVI). This latter is emulated by a bounded-size subset of vehicles currently populating the geographic region where the virtual infrastructure is to be deployed. An analytical model is proposed to study this mechanism. More precisely, this model is proposed to study the GVI in the frame of its main use: data dissemination in VANETs. Despite being simple, the proposed model can accurately predict the system performance such as the probability that a vehicle is informed, and the average number of duplicate messages received by a vehicle, and allows a careful investigation of the impact of vehicular traffic properties and system parameters on performance criteria. Analytical and simulation results show that the proposed GVI mechanism can periodically disseminate the data within an intersection area, efficiently utilize the limited bandwidth and ensure high delivery ratio.