CVI: Connected Vehicle Infrastructure for ITS

Intelligent Transportation Systems (ITS) aggregate, analyze and display geographic- and temporal-specific sensor information to reduce congestion while promoting safety. Until now, the coverage and potential of ITS have been restricted by the excessive cost of deploying the required road sensor and communications infrastructure. Our solution to this problem is focused on a novel integrated ITS Network Architecture where vehicles are the main infrastructure in the network. We propose The Connected Vehicle Infrastructure (CVI) for Intelligent Transportation Systems (ITS), where vehicles adapt to mobility changes to form stable vehicular clusters using a Network Criticality-based algorithm we have developed. Further, they build on their clusters to form more stable Mobile Networks, as part of the ITS network. Simulation results using NS-2 show that CVI clustering provides more stable clusters, lower handoffs, higher resilience to errors and better connectivity than popular density-based vehicle clustering methods. In addition, the overhead analysis of CVI shows that it achieves reasonable overhead compared to common clustering algorithms.