Non-recurrent classification learning Model for Drone Assisted Vehicular Ad-Hoc Network Communication in Smart Cities

Vehicular Ad-Hoc Network (VANET) relies on momentary and stable infrastructure units for seamless connectivity. A stabilized connection in the network is required to aid uninterrupted service support regardless of the frequent displacement in the vehicle position. The implication of drones or flying ad-hoc networks in VANETs is common in smart cities to provide transitory solutions for connectivity and data transfer. Considering the current and future applications of drones in diverse smart city applications, this article introduces the Decisive Data Dissemination Model (D3M). This proposed model is used for providing latency-less data transfer in drone-assisted VANET communications. The uncertainties in neighbor discovery due to handoff failure and access unavailability are addressed using non-recurrent classification learning. By using this learning, the splitting instance for handoff and neighbor requiring vehicles are identified to restore the fading connectivity. The re-connecting instances in the active and dissemination time are instantaneous using available drones and neighbors, preventing handoff failure. The performance of the proposed model is verified using simulations and it is seen that it achieves high data transfer rate, with less dissemination delay and backlogs.