Efficient Data Broadcast for Generic Applications Using a Scalable Dynamic VANET Algorithm

Data broadcasting using VANETs has been studied extensively. It is a very important part of applications of connected vehicles where there is a constant need for vehicles to exchange information with the communication infrastructure. This could result in the congestion of the network with data. Alternatively information can be disseminated using exclusively adhoc V2V approach. This approach may lead to broadcast storm and jamming of the V2V channel. In this study we evaluate as a proof-of-concept, V2V data broadcast, for an algorithm proposed in an earlier study to build autonomous and dynamic groups. In our recent studies we showed that the hybrid VANETs formed using this algorithm can reduce communication load and are very scalable for sensor data collection. In this study we modify the dynamic grouping algorithm so that the time for which the group is available for data broadcast is improved. This modified algorithm for dynamic grouping of vehicular nodes is implemented in a realistic well-calibrated microscopic traffic simulation test bed of New Jersey Turnpike. We demonstrate the benefit by considering a generic instance of an event-specific data broadcast. This data could be information on accidents, local commercial information, etc. We demonstrate, using initial findings, that the algorithm can be used to contain broadcast storm and reduce bandwidth usage by 65-76% compared to flooding and 45-62% compared to farthest neighbor broadcast protocol. Thus we establish that the dynamic grouping algorithm is an effective solution in many real-world traffic scenarios.